Skip to content

fix(kas): relay an MCP server's boolean env flags, never its secrets - #25331

Merged
buluoray merged 1 commit into
mainfrom
fix/kas-mcp-env-settings-25318
Oct 28, 2026
Merged

buluoray merged 1 commit into
mainfrom
fix/kas-mcp-env-settings-25318

Conversation

@buluoray

@buluoray buluoray commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: An MCP server declared in an agent spec starts on KAS with the same settings it gets on kiro-cli, while no credential-shaped env value reaches the ACP wire.

On KAS the agent spec travels over ACP (_meta.kiro.customAgents on session/new), so the projection in src/kiro_crew/acp/kas_agents.py withheld every literal env value from a server Crew did not author. A server that reads feature flags or settings from env started without them: in a live comparison one server exposed 15 tools on v2 and 37 on KAS from the same spec, and a second lost two settings variables.

Why it matters

kiro-cli moves its default ACP engine to KAS on 11/2, so every session that ran on v2 runs on KAS. Any agent-spec server configured through env silently changes behaviour on that date, with no error.

Not a goal

  • Relaying credentials to a server. A secret still has to reach a server another way (a secret:// reference routed through the gateway, or an edition-relayed ${NAME}).
  • Changing Crew's own managed servers. They still relay only KIROCREW_HOME (and edition references).
  • headers, command, args and url: unchanged.
  • Any non-boolean env value (words, numbers, URLs, paths). Those values stay withheld on KAS here and are named in the log. KAS: relay URL and path env values for agent-spec MCP servers via a local launcher #25362 owns relaying them through a local launcher so the env never crosses the wire.

What changed (motivation → approach → change)

Symptom: env settings missing on KAS. Root cause: _withhold_credential_fields kept only an exact edition-relayed ${NAME} on a third-party server and dropped every literal. It did this because anything kept is in the session/new frame, its recordings and its logs.

Approaches weighed:

  1. Launch such servers through the session-level hoist (hoist_servers_for_session) so env stays local. Rejected: the hoist moves an entry into session/new's own mcpServers array, which is the same frame, and hoists it as already projected. Keeping env truly local needs a Crew-side launcher, a new launch path; KAS: relay URL and path env values for agent-spec MCP servers via a local launcher #25362 owns that.
  2. Relay only literals judged non-secret. Chosen, then narrowed over seven review rounds. Every rule that relayed plain words or wider shapes behind key-name and value-shape checks met a credential name no list covered: camelCase keys, driver URLs, quoted JSON keys, a webhook token in a URL path, MYSQL_PWD, run-together DBPASS, acronym-prefixed names. So the rule now holds by construction: no secret is a boolean.

The change: _relayable_literal(key, value) relays a third-party literal only when it is exactly true or false (case-insensitive, trimmed; _RELAYABLE_BOOLEANS). The value is relayed as written. A boolean is still withheld when:

  • its key matches the shared config.sections_agent_env.AGENT_ENV_CREDENTIAL_CLASS (the name split at punctuation and camelCase), or the gateway's is_secret_env_key / is_credential_env_key prefixes;
  • the credential redactor (external_text.redact_external_text) would change it alone or beside its key.

Every other literal is withheld: numbers (a PIN is digits), words, URLs, paths, JSON, references, secret://. The projection's INFO line names a withheld key when the name is a plain word the redactor leaves alone, counts the rest, and never prints a value. Managed servers keep their existing rule. agent.env and the shared class are unchanged. acp/kas_agents.py is listed in security_posture.NON_EGRESS_REDACTION_MODULES: it asks the redactor a yes/no question and emits no redacted text.

This fixes the boolean feature-flag case the issue reports and leaves the rest to #25362, so the PR says Refs #25318 rather than closing it.

Backwards compatibility

Compatible: the change only relays values that were withheld before; nothing that was relayed is now dropped.

Removes nothing: no field, key, record or path is taken away; the only behaviour change is more env keys reaching third-party servers.

Tests

  • test/test_kas_agents_mcp_env_relay.py (new):
    • TestBooleansReachTheServer:
      • true / false in any case, and with surrounding spaces, relay as written. They appear in the serialized session/new params on both the new path and the load-path hoist.
      • A secret beside them costs only itself.
      • A managed server still relays no literal.
    • TestEverythingElseIsWithheld: 36 non-boolean values are withheld on both paths. Each case asserts its key is absent from the serialized params, and a value long enough to be unique is absent too.
      • Settings: LOG_LEVEL=debug, numbers, yes / on / 1, a host, a URL, a path.
      • Every short secret earlier rounds found under names no list covered: DBPASS, SMTPPASS, MYSQLPWD, ADMINPW, DB_PASSPHRASE, AWS_CREDS, SID_COOKIE, BASIC_AUTH, MYSQL_PWD, dbPass, SERVICE_PW, MODE, a digit PIN.
      • Credential shapes: ghp_, sk-proj-, AKIA, an opaque token, JSON, a connection string, a driver URL with a password, a webhook URL, ${}, secret://.
      • true;rm and a quoted "true".
    • TestABooleanCanStillBeWithheld: true under API_KEY, GITHUB_TOKEN, CLIENT_SECRET, DB_PASSWORD or encryptionKey is withheld, and so is a key-paired ghp_...=true.
    • TestTheRuleIsLocalToTheKasProjection: PW_TEST_HTML_REPORT_OPEN is withheld on KAS, while agent.env still saves and loads it.
    • TestTheWithholdIsLogged: withheld keys are named, values never are, and a credential-shaped key name is counted, not printed.
  • test/test_kas_agents_mcp_servers.py: three tests used a fake secret (sekrit, sk-live, n0nce) under an innocent key. They now put it under a credential key.

Mutation check on the pushed head. Each check was removed alone, one at a time, with the tree restored and clean in between:

Mutation Reddens
relay off (base behaviour) test_boolean_flags_reach_a_third_party_server and 80 more
boolean check off test_a_non_boolean_value_is_not_in_the_frame[LOG_LEVEL-*], [MAX_RESULTS-*] and 63 more
no trim / case fold test_each_boolean_relays_on_its_own[DEBUG], [TELEMETRY], [VERBOSE] and more
key check off test_a_boolean_under_a_credential_key_is_withheld[GITHUB_TOKEN], [CLIENT_SECRET], [DB_PASSWORD], [encryptionKey]
redactor check off test_a_key_paired_credential_is_withheld
key names not logged test_withheld_keys_are_named_and_values_are_not
key names unfiltered test_a_credential_shaped_key_name_is_counted_not_printed

Redundancy, by design: true under API_KEY is caught by both the key check and the redactor's key-anchored pattern.

Proof: PROVEN test/test_kas_agents_mcp_env_relay.py::TestBooleansReachTheServer::test_boolean_flags_reach_a_third_party_server, test/test_kas_agents_mcp_env_relay.py::TestBooleansReachTheServer::test_each_boolean_relays_on_its_own[DEBUG], test/test_kas_agents_mcp_env_relay.py::TestBooleansReachTheServer::test_each_boolean_relays_on_its_own[ENABLE_EXPERIMENTAL_TOOLS], test/test_kas_agents_mcp_env_relay.py::TestBooleansReachTheServer::test_each_boolean_relays_on_its_own[FEATURE_X_ENABLED], test/test_kas_agents_mcp_env_relay.py::TestBooleansReachTheServer::test_each_boolean_relays_on_its_own[TELEMETRY] (head 22b1554)

Locally run: test/test_kas_agents_mcp_env_relay.py, test/test_kas_agents_mcp_servers.py, test/test_kas_agents.py, test/test_security_posture.py, test/test_acp_agent_env.py and test/test_agent_spec_fields_doc.py, 725 tests, all pass on this head. test_every_backend_launch_carries_agent_env[copilot] was deselected: it is red on origin/main on this host too, because a .mcp.json in the home directory makes Copilot refuse to start. The full suite was not run locally.

Manual verification

Not run against a live KAS session on this host. Unit coverage asserts on the serialized session/new params, which is the frame KAS receives and what a frame recording stores. Windows and macOS are unverified here, though the change is pure string handling with no platform branch.

Related Issues

Refs #25318: the boolean feature-flag case.

#25362 owns relaying every other env value through a local launcher, so env never crosses the wire.

Sibling sweep

Swept: git grep -nE "_CREDENTIAL_BEARING_FIELDS|withhold.*env|env.*withheld" -- src/kiro_crew/acp
Sibling: src/kiro_crew/acp/session_mcp.py:183 -- not affected -- mirrored hosts' acp_server_element carries the spec env as given, so there is no withhold to narrow
Sibling: src/kiro_crew/acp/kas_agents.py:1437 -- not affected -- hoist_servers_for_session hoists the already projected entry, so it inherits this filter

Pattern harvest

Not generalizable: the one place an agent spec's env is filtered for a wire is this projection; the reusable pieces, the redactor and the shared credential-name class, are reused rather than respelled. The lesson recorded for the class: a value gate for a wire is an allowlist of shapes, never a denylist of credential forms.

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title (feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)
  • Existing tests pass and new tests added for new functionality
  • Self-review completed; code follows project style guidelines
  • Documentation updated (if applicable)
  • No secrets, credentials, or internal references in the diff

Contribution License Agreement

Contributed under the project's Apache-2.0 license.

@buluoray
buluoray requested review from a team and KiroCrewCaptain as code owners October 28, 2026 17:18
@buluoray
buluoray requested a review from cruisercohen October 28, 2026 17:18
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 28, 2026
@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ⏭️ skipped

Revision 22b1554893c927bb525300da19205357ecc29155 touches no user-facing surface (no changes under website/ or committed screenshots), so the UX review was skipped. Does not block merge.

@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ human override accepted

@buluoray overrode this lane for 22b1554893c927bb525300da19205357ecc29155 via /ai-review override (this commit only).

@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ✅ human override accepted

@buluoray overrode this lane for 22b1554893c927bb525300da19205357ecc29155 via /ai-review override (this commit only).

@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed 22b1554893c927bb525300da19205357ecc29155 — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] 22b1554

Verdict parsed from the review's SHA-scoped output markers for commit 22b1554893c927bb525300da19205357ecc29155.

False positive or not applicable? A repository writer can comment:
/ai-review override fable 22b1554893c927bb525300da19205357ecc29155 false-positive: <reason, 20 to 500 characters>

@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ human override accepted

Human judgment by @buluoray overrides the GPT 6.1 finding for 22b1554893c927bb525300da19205357ecc29155; the recorded reason is authoritative for this commit.

[GPT-OVERRIDE] 22b1554

This comment is updated in place on each push.

The model was not re-run because an authorized human decision supersedes it.

False positive or not applicable? A repository writer can comment:
/ai-review override gpt 22b1554893c927bb525300da19205357ecc29155 false-positive: <reason, 20 to 500 characters>

@buluoray
buluoray force-pushed the fix/kas-mcp-env-settings-25318 branch from 6e8357a to b3ef92b Compare October 28, 2026 17:37
@github-actions

github-actions Bot commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

Sensitive Change Review

Sensitive files changed:

  • src/kiro_crew/security_posture.py

  • @buluoray posted a complete reasoning template for 22b1554.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: dispositions for the review of 6e8357acfc, fixed at b3ef92bf56.

  • Opus 5.5, BLOCKING (camelCase / run-together credential keys). Fixed in code. Key names are now split at _, punctuation and camelCase boundaries before matching, and a credential word counts at the end of any segment, so accessKey, signingKey, encryptionKey, ENCRYPTIONKEY, dbPass and authTokenValue are withheld. Pinned by the [camel-*] cases in test/test_kas_agents_mcp_env_relay.py; removing the camelCase split reddens [camel-authTokenValue-*], and removing the extra segment words reddens [camel-dbPass-*].
  • GPT 6.1, BLOCKING (password in a driver URL). Fixed in code. Any scheme://user:pass@ is withheld regardless of scheme, so postgresql+asyncpg:// and mssql+pyodbc:// no longer depend on the redactor's scheme list. A value carrying an embedded name= / name: whose name is credential-shaped (Server=db;Pwd=…) is withheld as well. Pinned by [userinfo-*] and [embedded-*]; each reddens when its gate is removed.
  • First Principles note, item 4 (second spelling of the credential-name check). Taken. The key check now applies the shared AGENT_ENV_CREDENTIAL_CLASS first and adds only the short spellings and bearer words that class leaves out, with the same segment-end anchoring. GOPRIVATE, TOKENIZERS_PARALLELISM and AUTHOR_NAME now relay, and a test pins that.
  • First Principles item 3 / Design note (naming withheld keys in the log). Taken. The key naming is dropped and the documented rule is restored: the INFO line names the server and field, never a key or value.
  • Static Ratchet Gate (test_every_redactor_call_site_is_a_registered_sink_or_allowlisted). Fixed: acp/kas_agents.py is listed in NON_EGRESS_REDACTION_MODULES with its reason (it asks the redactor a yes/no question and emits no redacted text).

The residual stays as stated in security.md: a secret with no credential shape under an innocent key name is still relayed, because the gate is shape-based.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: sensitive change review for the current head.

Sensitive change review

  • Head commit: b3ef92b
  • What rule changed: security_posture.NON_EGRESS_REDACTION_MODULES gains one entry, acp/kas_agents.py, with its reason. That module now calls external_text.redact_external_text as a yes/no predicate inside _relayable_literal, to decide whether a third-party MCP server's literal env value may go into the KAS session/new frame. No sink, pattern, path rule or redaction behaviour changes.
  • Before -> after (what was allowed/blocked, now): Before, the drift guard test_every_redactor_call_site_is_a_registered_sink_or_allowlisted failed because acp/kas_agents.py called a redactor without being a sink or allowlisted. After, it is allowlisted as gate-side: the redactor's answer withholds a value; nothing the module emits is redacted text. The Security panel's redaction_paths count is unchanged, which is correct because no new egress path was added.
  • Worst case if wrong (breaks commands / leaks secret / locks user out): if the module were really an egress sink, the panel would under-count egress paths by one. No command breaks, no secret is exposed and nobody is locked out by this entry; the module's actual withholding is pinned separately by test/test_kas_agents_mcp_env_relay.py.
  • Evidence checked (which sandbox result proves it): test/test_security_posture.py passes with the entry and failed in CI on 6e8357acfc without it. kas_agents.py was read to confirm it returns only the redactor's yes/no answer: the INFO log names the server and field, never a key or value, and test_the_server_is_named_and_neither_key_nor_value_is pins that.
  • No regression:
    • Tools still work (which tools/commands were run, result): pytest -n0 on test_kas_agents_mcp_env_relay.py, test_kas_agents_mcp_servers.py, test_kas_agents.py, test_security_posture.py, test_agent_spec_fields_doc.py: 404 passed. flake8 and black clean on the changed Python files.
    • Backward compatible (old config / old data / old callers still work, how checked): the change adds one string to a frozenset read only by the posture registry and its drift test. No config key, data format or public function changed.
    • Existing tests (which suites passed on this head): the targeted suites above locally; CI on this head is still running at the time of posting.
  • How to undo: remove the acp/kas_agents.py entry from NON_EGRESS_REDACTION_MODULES in src/kiro_crew/security_posture.py, or revert commit b3ef92b.

@buluoray
buluoray force-pushed the fix/kas-mcp-env-settings-25318 branch from b3ef92b to 453177f Compare October 28, 2026 18:03
@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: dispositions for the review of b3ef92bf56, addressed at 453177fd0c.

  • Opus 5.5, BLOCKING (a quoted JSON key inside a value is relayed). Fixed in code by changing the approach. This was the third finding of one class (a credential form the value denylist did not know), so the value check is now an allowlist of shapes: a plain word ([A-Za-z0-9._-]), an http(s) URL with no userinfo, query or fragment, or an absolute path. JSON, k=v lists, connection strings, quoted text, $ references and secret:// are withheld without having to be recognised. The JSON blob (two quote styles), k=v, connection-string and quoted-string cases are in _CREDENTIALS; removing the shape allowlist reddens 24 shape cases.
  • GPT 6.1, BLOCKING on 6e8357acfc (password in a driver URL). Still covered: a URL with user:pass@ is not an allowed shape under any scheme, so the https, postgresql+asyncpg and mssql+pyodbc cases are withheld by the allowlist itself.
  • First Principles notes (local userinfo regex; KAS-only word list). Taken. Both are deleted. The key check is now only the shared AGENT_ENV_CREDENTIAL_CLASS (with camelCase split) plus the gateway's prefixes. The redactor and opaque-token checks stay on plain words, because a bare ghp_ / sk- / AKIA token is itself a plain word.
  • Design notes. (1) SESSION_TIMEOUT, AUTH_MODE, PIN_MEMORY and the like now relay, since the KAS-only words are gone. (2) Fixtures whose fake secret sat under an innocent key were moved under credential keys; the withhold-by-shape behaviour is now covered directly by the shape cases.

Residual, stated in security.md and pinned by TestResiduals: a short plain-word secret under a key the shared class does not know (dbPass=hunter3, SERVICE_PW=letmein) is relayed.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: sensitive change review for the current head.

Sensitive change review

  • Head commit: 453177f
  • What rule changed: security_posture.NON_EGRESS_REDACTION_MODULES gains one entry, acp/kas_agents.py, with its reason. That module calls external_text.redact_external_text only as a yes/no predicate: in _relayable_literal, to decide whether a third-party MCP server's literal env value may go into the KAS session/new frame, and in _loggable_key_names, to decide whether a withheld key NAME may be printed in the log as written. No sink, pattern, path rule or redaction behaviour changes.
  • Before -> after (what was allowed/blocked, now): Before, the drift guard test_every_redactor_call_site_is_a_registered_sink_or_allowlisted failed because acp/kas_agents.py called a redactor without being a sink or allowlisted. After, it is allowlisted as gate-side. It never emits redacted text: a name the redactor would change is counted as "N unnamed", not printed. The Security panel's redaction_paths count is unchanged.
  • Worst case if wrong (breaks commands / leaks secret / locks user out): if the module were really an egress sink, the panel would under-count egress paths by one. No command breaks, no secret is exposed and nobody is locked out by this entry. The module's withholding is pinned separately by test/test_kas_agents_mcp_env_relay.py.
  • Evidence checked (which sandbox result proves it): test/test_security_posture.py passes with the entry, and failed in CI on 6e8357acfc without it. test_a_credential_shaped_key_name_is_counted_not_printed shows a ghp_-shaped key name is counted and never printed; removing that filter reddens it.
  • No regression:
    • Tools still work (which tools/commands were run, result): pytest -n0 on test_kas_agents_mcp_env_relay.py, test_kas_agents_mcp_servers.py, test_kas_agents.py, test_security_posture.py, test_agent_spec_fields_doc.py: 422 passed. flake8 and black are clean on the changed Python files.
    • Backward compatible (old config / old data / old callers still work, how checked): the change adds one string to a frozenset read only by the posture registry and its drift test. No config key, data format or public function changed.
    • Existing tests (which suites passed on this head): the targeted suites above passed locally. CI on this head is still running at the time of posting.
  • How to undo: remove the acp/kas_agents.py entry from NON_EGRESS_REDACTION_MODULES in src/kiro_crew/security_posture.py, or revert commit 453177f.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: dispositions for the review of 453177fd0c, addressed at 439183872b.

  • GPT 6.1, BLOCKING (a webhook token in a URL path is relayed). Fixed in code by narrowing the scope. The URL and path shapes are gone, and only plain words ([A-Za-z0-9._-], bools and numbers included) can cross the wire. Every URL and path is now withheld and named in the projection log. The webhook URL, plain base and local URLs, and POSIX and Windows paths are all in _CREDENTIALS as shape cases. Removing the plain-word shape check reddens all 36 shape tests.
  • URL and path env on KAS are tracked in KAS: relay URL and path env values for agent-spec MCP servers via a local launcher #25362, which proposes a local launcher so env never crosses the wire. The PR body's "Not a goal" section and the spec (acp-client.md, security.md, agent-spec-fields.md) both link it.
  • Earlier findings. The camelCase keys (Opus), the driver-URL password (GPT) and the quoted JSON key (Opus) are each still withheld, because none of those values is a plain word. The key check covers DB_PASSWORD=hunter2 and API_KEY=abc123, and the redactor covers bare ghp_ / sk-proj- / AKIA tokens.

Residual, stated in security.md and pinned by TestResiduals: a short plain-word secret under a key the shared credential class does not know (dbPass=hunter3) is relayed.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: sensitive change review for the current head.

Sensitive change review

  • Head commit: 4391838
  • What rule changed: security_posture.NON_EGRESS_REDACTION_MODULES gains one entry, acp/kas_agents.py, with its reason. That module calls external_text.redact_external_text only as a yes/no predicate: in _relayable_literal, to decide whether a third-party MCP server's literal env value may go into the KAS session/new frame, and in _loggable_key_names, to decide whether a withheld key NAME may be printed in the log as written. No sink, pattern, path rule or redaction behaviour changes.
  • Before -> after (what was allowed/blocked, now): Before, the drift guard test_every_redactor_call_site_is_a_registered_sink_or_allowlisted failed because acp/kas_agents.py called a redactor without being a sink or allowlisted. After, it is allowlisted as gate-side. It never emits redacted text: a name the redactor would change is counted as "N unnamed", not printed. The Security panel's redaction_paths count is unchanged.
  • Worst case if wrong (breaks commands / leaks secret / locks user out): if the module were really an egress sink, the panel would under-count egress paths by one. No command breaks, no secret is exposed and nobody is locked out by this entry. The module's withholding is pinned separately by test/test_kas_agents_mcp_env_relay.py.
  • Evidence checked (which sandbox result proves it): test/test_security_posture.py passes with the entry, and failed in CI on 6e8357acfc without it. test_a_credential_shaped_key_name_is_counted_not_printed shows a ghp_-shaped key name is counted and never printed; removing that filter reddens it.
  • No regression:
    • Tools still work (which tools/commands were run, result): pytest -n0 on test_kas_agents_mcp_env_relay.py, test_kas_agents_mcp_servers.py, test_kas_agents.py, test_security_posture.py, test_agent_spec_fields_doc.py: 429 passed. flake8 and black are clean on the changed Python files.
    • Backward compatible (old config / old data / old callers still work, how checked): the change adds one string to a frozenset read only by the posture registry and its drift test. No config key, data format or public function changed.
    • Existing tests (which suites passed on this head): the targeted suites above passed locally. CI on this head is still running at the time of posting.
  • How to undo: remove the acp/kas_agents.py entry from NON_EGRESS_REDACTION_MODULES in src/kiro_crew/security_posture.py, or revert commit 4391838.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 28, 2026
@buluoray
buluoray force-pushed the fix/kas-mcp-env-settings-25318 branch from 4391838 to bc645bc Compare October 28, 2026 18:45
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention labels Oct 28, 2026
@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: disposition for the review of 439183872b, addressed at bc645bcf30.

  • GPT 6.1, BLOCKING (MYSQL_PWD / DB_PASS / SERVICE_PW short passwords relay). The owner chose to fix this in the shared class. config.sections_agent_env.AGENT_ENV_CREDENTIAL_CLASS now also matches PWD, PASS or PW as a whole segment, (^|_)(PWD|PASS|PW)(_|$), case-insensitive. The KAS projection splits camelCase first, so dbPass is covered too. MYSQL_PWD, DB_PASS, dbPass, SERVICE_PW, PW and smtp-pass are now withheld, while PASSTHROUGH, BYPASS_CACHE, COMPASS_MODE and PWDIR still relay. The former residual pin is now this withheld case.
  • The class's other consumer is agent.env. It now refuses the same names, and test/test_acp_agent_env.py gains both lists, accepted and refused. Two mutations were each run alone. Disabling the new alternative reddens 12 named tests across both files. Letting it match at the end of a longer segment reddens the BYPASS_CACHE, COMPASS_MODE and SPW_MODE acceptance tests.
  • Owning specs were updated in the same commit: config.md (agent.env refusal rule) and security.md (the agent.env property and the KAS threat-model row).
  • Design note (LAUNCH_NONCE, PASSPHRASE, PIN, SALT, COOKIE). These are not in the shared class, so as plain words they would relay. The remaining residual is a short plain-word secret under a name the class does not know, and the threat model says so. Per the Design lane's own note, this filter stays a stopgap, and KAS: relay URL and path env values for agent-spec MCP servers via a local launcher #25362 is the real fix.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: sensitive change review for the current head.

Sensitive change review

  • Head commit: bc645bc
  • What rule changed: security_posture.NON_EGRESS_REDACTION_MODULES gains one entry, acp/kas_agents.py, with its reason. That module calls external_text.redact_external_text only as a yes/no predicate: in _relayable_literal, to decide whether a third-party MCP server's literal env value may go into the KAS session/new frame, and in _loggable_key_names, to decide whether a withheld key NAME may be printed in the log as written. No sink, pattern, path rule or redaction behaviour changes.
  • Before -> after (what was allowed/blocked, now): Before, the drift guard test_every_redactor_call_site_is_a_registered_sink_or_allowlisted failed because acp/kas_agents.py called a redactor without being a sink or allowlisted. After, it is allowlisted as gate-side. It never emits redacted text: a name the redactor would change is counted as "N unnamed", not printed. The Security panel's redaction_paths count is unchanged.
  • Worst case if wrong (breaks commands / leaks secret / locks user out): if the module were really an egress sink, the panel would under-count egress paths by one. No command breaks, no secret is exposed and nobody is locked out by this entry. The module's withholding is pinned separately by test/test_kas_agents_mcp_env_relay.py.
  • Evidence checked (which sandbox result proves it): test/test_security_posture.py passes with the entry, and failed in CI on 6e8357acfc without it. test_a_credential_shaped_key_name_is_counted_not_printed shows a ghp_-shaped key name is counted and never printed; removing that filter reddens it.
  • No regression:
    • Tools still work (which tools/commands were run, result): pytest -n0 on test_kas_agents_mcp_env_relay.py, test_kas_agents_mcp_servers.py, test_kas_agents.py, test_security_posture.py, test_agent_spec_fields_doc.py, test_acp_agent_env.py, test_native_allow_stores.py: 766 passed (one host-dependent copilot launch test deselected; it is red on origin/main on this host too). flake8 and black are clean on the changed Python files.
    • Backward compatible (old config / old data / old callers still work, how checked): the change adds one string to a frozenset read only by the posture registry and its drift test. No config key, data format or public function changed.
    • Existing tests (which suites passed on this head): the targeted suites above passed locally. CI on this head is still running at the time of posting.
  • How to undo: remove the acp/kas_agents.py entry from NON_EGRESS_REDACTION_MODULES in src/kiro_crew/security_posture.py, or revert commit bc645bc.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 28, 2026
@buluoray
buluoray enabled auto-merge (squash) October 28, 2026 18:53
@buluoray
buluoray force-pushed the fix/kas-mcp-env-settings-25318 branch from 1970c2e to 22b1554 Compare October 28, 2026 19:55
@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: reply to the Opus 5.5 finding (run-together DBPASS, SMTPPASS, MYSQLPWD, ADMINPW) and the GPT 6.1 finding (acronym-prefixed credential names bypass the token split) on 1970c2e25b. Both are fixed in code at 22b1554893, by narrowing the rule. No name was added.

  • The rule is now boolean-only. A third-party server's literal env value crosses the KAS wire only when it is exactly true or false (case-insensitive, trimmed). No secret is a boolean, so the rule no longer depends on which key names or value shapes carry a credential. That dependency is what produced both findings. The token-prefix list is gone.
  • Two checks still run in front of the boolean. The shared AGENT_ENV_CREDENTIAL_CLASS key check and the credential redactor, so API_KEY=true and a key-paired ghp_...=true are withheld.
  • Every other literal is withheld and its key named in the log. That includes numbers, since a PIN is digits. Your names are now withheld test cases in TestEverythingElseIsWithheld, alongside the earlier rounds' names: DBPASS, SMTPPASS, MYSQLPWD, ADMINPW, DB_PASSPHRASE, AWS_CREDS, SID_COOKIE, BASIC_AUTH, MYSQL_PWD, dbPass, SERVICE_PW. Removing the boolean check reddens 65 of those cases.
  • Scope. The PR now says Refs #25318. KAS: relay URL and path env values for agent-spec MCP servers via a local launcher #25362 owns relaying every non-boolean value through a local launcher, so env never crosses the wire. agent.env and the shared class are unchanged.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: sensitive change review for the current head.

Sensitive change review

  • Head commit: 22b1554
  • What rule changed: security_posture.NON_EGRESS_REDACTION_MODULES gains one entry, acp/kas_agents.py, with its reason. That module calls external_text.redact_external_text only as a yes/no predicate: in _relayable_literal, to decide whether a third-party MCP server's literal env value may go into the KAS session/new frame, and in _loggable_key_names, to decide whether a withheld key NAME may be printed in the log as written. No sink, pattern, path rule or redaction behaviour changes.
  • Before -> after (what was allowed/blocked, now): Before, the drift guard test_every_redactor_call_site_is_a_registered_sink_or_allowlisted failed because acp/kas_agents.py called a redactor without being a sink or allowlisted. After, it is allowlisted as gate-side. It never emits redacted text: a name the redactor would change is counted as "N unnamed", not printed. The Security panel's redaction_paths count is unchanged.
  • Worst case if wrong (breaks commands / leaks secret / locks user out): if the module were really an egress sink, the panel would under-count egress paths by one. No command breaks, no secret is exposed and nobody is locked out by this entry. The module's withholding is pinned separately by test/test_kas_agents_mcp_env_relay.py.
  • Evidence checked (which sandbox result proves it): test/test_security_posture.py passes with the entry, and failed in CI on 6e8357acfc without it. test_a_credential_shaped_key_name_is_counted_not_printed shows a ghp_-shaped key name is counted and never printed; removing that filter reddens it.
  • No regression:
    • Tools still work (which tools/commands were run, result): pytest -n0 on test_kas_agents_mcp_env_relay.py, test_kas_agents_mcp_servers.py, test_kas_agents.py, test_security_posture.py, test_agent_spec_fields_doc.py, test_acp_agent_env.py, test_native_allow_stores.py: 725 passed (one host-dependent copilot launch test deselected; it is red on origin/main on this host too). flake8 and black are clean on the changed Python files.
    • Backward compatible (old config / old data / old callers still work, how checked): the change adds one string to a frozenset read only by the posture registry and its drift test. No config key, data format or public function changed.
    • Existing tests (which suites passed on this head): the targeted suites above passed locally. CI on this head is still running at the time of posting.
  • How to undo: remove the acp/kas_agents.py entry from NON_EGRESS_REDACTION_MODULES in src/kiro_crew/security_posture.py, or revert commit 22b1554.

@buluoray buluoray changed the title fix(kas): relay an MCP server's env settings, never its secrets fix(kas): relay an MCP server's boolean env flags, never its secrets Oct 28, 2026
@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: disposition for the Design and First Principles BLOCKs on 22b1554893 ("the description documents a different rule than the diff ships").

Both lanes ran at 19:56Z and read the description from before this head. The description was rewritten at 19:59Z to match the bool-only rule:

Both jobs have been re-run against the current description.

One sentence is still out of date: the Goal line ("the same settings it gets on kiro-cli"). That section is frozen for agents, so I have not edited it. Its scope is now a boolean flag only, and the "Not a goal" section says so explicitly. The owner can reword the Goal if they want it to match.

@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: disposition for the First Principles note on 22b1554893 (item 2: drop the key-name check in front of a boolean).

Kept, by owner ruling. The ruling keeps the shared AGENT_ENV_CREDENTIAL_CLASS check and the redactor in front of the boolean rule. The key check guards nothing secret, and the note is right about that. It is kept as a conservative choice: the owner decided that a flag under a credential-shaped name (*_KEY, *_TOKEN) is withheld rather than relayed. The cost the note names is accepted. PW_TEST_HTML_REPORT_OPEN is not affected, since the shared class does not match it. true under API_KEY is withheld and logged by name, and #25362 relays it through the local launcher.

The GPT 6.1 Review check and the Sensitive Change Review check failed on this head because GitHub's primary API rate limit was hit. That is not a lane defect. Each will be re-run once after the hourly reset.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 28, 2026
@buluoray

Copy link
Copy Markdown
Contributor Author

🤖 Kiro Crew Auto-Pipeline [operator: buluoray#ce473fca]: Owner decision, posted at @buluoray's direction.

I accept the residual GPT 6.1 flagged on head 22b1554. A value is relayed to KAS only when it is exactly true or false, and the shared credential key check and redactor still run first. The only way a secret crosses the wire is if someone's password is literally true or false, which leaks at most one bit. That is not a real-world setup, and it does not justify withholding every env feature flag on KAS before kiro-cli makes v3 its default on 11/2.

Every other value stays withheld and is named in the log. #25362 (local launcher, env never on the wire) owns the full fix, including this residual. Releasing the GPT lane as the owner.

@buluoray

Copy link
Copy Markdown
Contributor Author

/ai-review override gpt 22b1554 owner-decision: Bool-only relay; the residual is a password literally true/false (at most one bit). Full fix tracked in #25362. Owner decision: #25331 (comment)

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@buluoray recorded a owner-decision release (class owner-decision) for the gpt AI finding on 22b1554893c927bb525300da19205357ecc29155.

Bool-only relay; the residual is a password literally true/false (at most one bit). Full fix tracked in #25362. Owner decision: #25331 (comment)

This decision applies only to this commit. A new push requires a new judgment.

@buluoray

Copy link
Copy Markdown
Contributor Author

/ai-review override design 22b1554 evidence-supplied: The earlier BLOCK sample read a stale PR description; the re-run on the current body passes: https://github.com/kirodotdev/KiroCrew/actions/runs/40758711702/job/123072972813

@buluoray

Copy link
Copy Markdown
Contributor Author

/ai-review override first-principles 22b1554 evidence-supplied: The earlier BLOCK sample read a stale PR description; the re-run on the current body passes: https://github.com/kirodotdev/KiroCrew/actions/runs/40758711830/job/123072982553

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@buluoray recorded a evidence-supplied release (class evidence-supplied) for the design AI finding on 22b1554893c927bb525300da19205357ecc29155.

The earlier BLOCK sample read a stale PR description; the re-run on the current body passes: https://github.com/kirodotdev/KiroCrew/actions/runs/40758711702/job/123072972813

This decision applies only to this commit. A new push requires a new judgment.

@github-actions

Copy link
Copy Markdown
Contributor

Human judgment recorded

@buluoray recorded a evidence-supplied release (class evidence-supplied) for the first-principles AI finding on 22b1554893c927bb525300da19205357ecc29155.

The earlier BLOCK sample read a stale PR description; the re-run on the current body passes: https://github.com/kirodotdev/KiroCrew/actions/runs/40758711830/job/123072982553

This decision applies only to this commit. A new push requires a new judgment.

@github-actions github-actions Bot added readiness: passed Eligible automated validation passed for the current revision and removed readiness: action required A blocking check or review needs attention labels Oct 28, 2026

@CrysisDeu CrysisDeu left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approve at 22b15548. Maintainer sign-off on the three self-overrides included below.

Rule as shipped: a third-party server's env value crosses the KAS wire only when it is the string true or false (case and edge whitespace ignored), the key is not credential-shaped, and the redactor leaves it alone. Everything else is withheld, so it fails closed. We ran 33 adversarial values through the real function (on, yes, 1, 0x1, secret://, ${}, empty, token-shaped keys): every result matched the rule. Residual leak is one bit (a password that is literally true/false). KAS gets strictly less than kiro-cli reads from disk. Sensitive Change Review and Security Scope green with the template for this head; security.md, acp-client.md, agent-spec-fields.md updated.

Self-overrides on this head, accepted as maintainer: GPT (owner decision on the one-bit residual: accepted); Design and First Principles (they read the stale body; re-runs on the current body pass). Next time ping a maintainer instead of self-signing; see #24692.

Nits: withheld key names now appear in INFO logs, including opaque ones, and main never logged them on purpose, so drop or downgrade that line; normalise the boolean before relay rather than sending whitespace as written; decide whether JSON true (not a string) should count, and say so in the docs.

Posted by code-reviewer (Kiro Crew member) on behalf of @CrysisDeu. Full report on file.

@buluoray
buluoray merged commit b734a68 into main Oct 28, 2026
117 of 130 checks passed
@buluoray
buluoray deleted the fix/kas-mcp-env-settings-25318 branch October 28, 2026 22:45
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Oct 28, 2026
bolichen97 pushed a commit that referenced this pull request Oct 31, 2026
Backports these main PRs to release/0.11.0, in pick order. Each was a
`git cherry-pick -x` of its squash commit on main:

- fix(memory): read memory.db off the writer lock, split background pool (#24290) (cherry picked from commit fed830c)
- fix(dashboard): bound the pre-prompt build's wait to enter the memory pool (#24414) (cherry picked from commit 520cf2c)
- fix(memory): keep the shared -wal when another process closes the memory store (#25852) (cherry picked from commit dc63bb6)
- fix: read the page-cache prefetch in a child to keep sqlite locks (#26302) (cherry picked from commit 81dde23)
- perf(memory): score facts from a term index and prebuild after restart (#25328) (cherry picked from commit 6824348)
- fix(desktop): retry a failed netstat before Windows recovery gives up (#25792) (cherry picked from commit 043dd9f)
- fix(cli): say a gateway still serves the port when restart's replacement is refused (#25366) (cherry picked from commit 2e13127)
- fix(cli): refuse to stop or restart a gateway the desktop app supervises (#25705) (cherry picked from commit bbf0910)
- fix(hooks): approve kiro-cli read-only file searches after the deny tiers (#25150) (cherry picked from commit a0c0bd4)
- fix(settings): refuse a config save that config.local.json overrides (#26216) (cherry picked from commit 7f8688e)
- fix(workflows): write non-finite floats as null so runs open (#26085) (cherry picked from commit 3591282)
- fix(approvals): give each background tool approval its own id (#25905) (cherry picked from commit 51ef3bb)
- fix(chat): record file changes for kiro-cli v3 write tools (#25650) (cherry picked from commit 5369270)
- fix(sandbox): stop blaming a host-wide OOM kill on a per-tree limit (#25647) (cherry picked from commit 791d2a2)
- fix(acp): replay history when a KAS resume restores no turns (#25521) (cherry picked from commit 47c71cb)
- fix(kas): auto-approve granted built-ins after Crew's own checks (#25339) (cherry picked from commit dba4de5)
- fix(kas): mount web_search as the engine's remote_web_search (#26435) (cherry picked from commit 65c21b5)
- fix(kas): answer a spec's shell.allowedCommands at the gate, after Crew's checks (#25539) (cherry picked from commit 39d07d9)
- fix(kas): relay an MCP server's boolean env flags, never its secrets (#25331) (cherry picked from commit b734a68)
- fix(security): read the -c script past options and redirections (#26087) (cherry picked from commit fea0920)
- fix(kas): refuse a shell command a denyByDefault spec does not allow (#26501) (cherry picked from commit 9f0d02f)
- fix(kas): name a deniedCommands deny aimed at a redirect as not carried (#26489) (cherry picked from commit 4951c0c)
- fix(security): delete the curl-upload deny rule (#26469) (cherry picked from commit 70e206e)

Conflicts resolved by hand:
- #26302: test/spawn_audit_helpers.py adds only the page_cache.py entry
  (the embeddings_child.py entry names a module 0.11.0 lacks);
  test/test_page_cache_prefetch.py takes main's file.
- #25328: test/test_dashboard_server_composition_contract.py lazy-import
  count 93 -> 96 (the three names it adds).
- #25705: test/test_cli_restart_refused_message.py takes main's file.
- #26085: dashboard/handlers/workflows.py adds import math, keeps the
  release typing import.
- #25339: platform/governance.py keeps the release rule for a missing
  ceiling (main's enterprise-baseline clause is not on 0.11.0).
- #26435: harness-parity.md H6 row edited in place on the release text.
- #25539: its acp-client.md and agent-spec-fields.md lines are left out;
  the paragraphs they edit are not on 0.11.0 (security.md carries it).
- #25331: the test and two doc mentions use hoist_managed_servers /
  resume=, the 0.11.0 names of main's renamed hoist_servers_for_session.
- #26087: security package line budget 32,552 -> 32,714 (re-measured).
- #26501: kas_permissions.py comment keeps 0.11.0's shapes (main's
  subagent rules are absent); its acp-client.md line is left out, and the
  agent-spec-fields.md toolsSettings row says denyByDefault is gate-enforced.
- #26489: its agent-spec-fields.md sentence is added to the 0.11.0 row.
- #26469: test_denied_commands_security.py keeps 0.11.0's comment; count 113 -> 112.

Test fixes (host-dependent tests made hermetic; none loosens an assertion):
- tmpdir_helpers.short_tmp_path: a tmp_path override for modules whose code
  judges a path by its components (a .kiro/crew data-home part); an agent
  session's TMPDIR under ~/.kiro/crew made every test path the data home.
  Used by the file-explorer, design-tweak (both), spawn-offload, launch-copy,
  home-pin, native-skill and auto-improvement-members tests; the sage
  ancestor-chain test creates its dir under /tmp for the same reason.
- pi-gate tests and the remote-session Python 3.9 test take the first real
  binary on PATH (playwright_cli_installer_helpers._native_tool_on_path): a
  mise shim under a scratch HOME hangs or writes to stdout.
- snapshot-merge ACL test names its own uid; nobody (65534) is the overflow
  uid inside a user namespace and cannot be named in an ACL.
- installer-step and push_guard interrupt tests spawn with SIGINT at its
  default; a background pytest run starts with SIGINT ignored, which a shell
  trap and Python both keep.

Left out: #23920 (already on release/0.11.0), #26144 (needs
sandbox_launch_mark from #24709), #26088 (fences a main-only port),
#25346 (11 of its tests fail on 0.11.0: the kiro-cli skill scope it
compares against differs there), #26397 (needs #26152's new download
route, not on 0.11.0), #26428 (not merged).
bolichen97 added a commit that referenced this pull request Oct 31, 2026
…26487)

Backports these main PRs to release/0.11.0, in pick order. Each was a
`git cherry-pick -x` of its squash commit on main:

- fix(memory): read memory.db off the writer lock, split background pool (#24290) (cherry picked from commit fed830c)
- fix(dashboard): bound the pre-prompt build's wait to enter the memory pool (#24414) (cherry picked from commit 520cf2c)
- fix(memory): keep the shared -wal when another process closes the memory store (#25852) (cherry picked from commit dc63bb6)
- fix: read the page-cache prefetch in a child to keep sqlite locks (#26302) (cherry picked from commit 81dde23)
- perf(memory): score facts from a term index and prebuild after restart (#25328) (cherry picked from commit 6824348)
- fix(desktop): retry a failed netstat before Windows recovery gives up (#25792) (cherry picked from commit 043dd9f)
- fix(cli): say a gateway still serves the port when restart's replacement is refused (#25366) (cherry picked from commit 2e13127)
- fix(cli): refuse to stop or restart a gateway the desktop app supervises (#25705) (cherry picked from commit bbf0910)
- fix(hooks): approve kiro-cli read-only file searches after the deny tiers (#25150) (cherry picked from commit a0c0bd4)
- fix(settings): refuse a config save that config.local.json overrides (#26216) (cherry picked from commit 7f8688e)
- fix(workflows): write non-finite floats as null so runs open (#26085) (cherry picked from commit 3591282)
- fix(approvals): give each background tool approval its own id (#25905) (cherry picked from commit 51ef3bb)
- fix(chat): record file changes for kiro-cli v3 write tools (#25650) (cherry picked from commit 5369270)
- fix(sandbox): stop blaming a host-wide OOM kill on a per-tree limit (#25647) (cherry picked from commit 791d2a2)
- fix(acp): replay history when a KAS resume restores no turns (#25521) (cherry picked from commit 47c71cb)
- fix(kas): auto-approve granted built-ins after Crew's own checks (#25339) (cherry picked from commit dba4de5)
- fix(kas): mount web_search as the engine's remote_web_search (#26435) (cherry picked from commit 65c21b5)
- fix(kas): answer a spec's shell.allowedCommands at the gate, after Crew's checks (#25539) (cherry picked from commit 39d07d9)
- fix(kas): relay an MCP server's boolean env flags, never its secrets (#25331) (cherry picked from commit b734a68)
- fix(security): read the -c script past options and redirections (#26087) (cherry picked from commit fea0920)
- fix(kas): refuse a shell command a denyByDefault spec does not allow (#26501) (cherry picked from commit 9f0d02f)
- fix(kas): name a deniedCommands deny aimed at a redirect as not carried (#26489) (cherry picked from commit 4951c0c)
- fix(security): delete the curl-upload deny rule (#26469) (cherry picked from commit 70e206e)

Conflicts resolved by hand:
- #26302: test/spawn_audit_helpers.py adds only the page_cache.py entry
  (the embeddings_child.py entry names a module 0.11.0 lacks);
  test/test_page_cache_prefetch.py takes main's file.
- #25328: test/test_dashboard_server_composition_contract.py lazy-import
  count 93 -> 96 (the three names it adds).
- #25705: test/test_cli_restart_refused_message.py takes main's file.
- #26085: dashboard/handlers/workflows.py adds import math, keeps the
  release typing import.
- #25339: platform/governance.py keeps the release rule for a missing
  ceiling (main's enterprise-baseline clause is not on 0.11.0).
- #26435: harness-parity.md H6 row edited in place on the release text.
- #25539: its acp-client.md and agent-spec-fields.md lines are left out;
  the paragraphs they edit are not on 0.11.0 (security.md carries it).
- #25331: the test and two doc mentions use hoist_managed_servers /
  resume=, the 0.11.0 names of main's renamed hoist_servers_for_session.
- #26087: security package line budget 32,552 -> 32,714 (re-measured).
- #26501: kas_permissions.py comment keeps 0.11.0's shapes (main's
  subagent rules are absent); its acp-client.md line is left out, and the
  agent-spec-fields.md toolsSettings row says denyByDefault is gate-enforced.
- #26489: its agent-spec-fields.md sentence is added to the 0.11.0 row.
- #26469: test_denied_commands_security.py keeps 0.11.0's comment; count 113 -> 112.

Test fixes (host-dependent tests made hermetic; none loosens an assertion):
- tmpdir_helpers.short_tmp_path: a tmp_path override for modules whose code
  judges a path by its components (a .kiro/crew data-home part); an agent
  session's TMPDIR under ~/.kiro/crew made every test path the data home.
  Used by the file-explorer, design-tweak (both), spawn-offload, launch-copy,
  home-pin, native-skill and auto-improvement-members tests; the sage
  ancestor-chain test creates its dir under /tmp for the same reason.
- pi-gate tests and the remote-session Python 3.9 test take the first real
  binary on PATH (playwright_cli_installer_helpers._native_tool_on_path): a
  mise shim under a scratch HOME hangs or writes to stdout.
- snapshot-merge ACL test names its own uid; nobody (65534) is the overflow
  uid inside a user namespace and cannot be named in an ACL.
- installer-step and push_guard interrupt tests spawn with SIGINT at its
  default; a background pytest run starts with SIGINT ignored, which a shell
  trap and Python both keep.

Left out: #23920 (already on release/0.11.0), #26144 (needs
sandbox_launch_mark from #24709), #26088 (fences a main-only port),
#25346 (11 of its tests fail on 0.11.0: the kiro-cli skill scope it
compares against differs there), #26397 (needs #26152's new download
route, not on 0.11.0), #26428 (not merged).

Co-authored-by: w <w@k.local>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants