feat(skills): Cursor-native flatten and 2026 high-signal stack - #42
Merged
Conversation
Cursor indexes one-level skill dirs (.cursor/skills/<name>/SKILL.md), so the nested .claude/skills/gstack/<skill> pack was invisible as /plan-ceo-review. Flatten each skill into .cursor/skills and ~/.cursor/skills, and vendor gstack bin/setup so the skill-start preamble is runnable. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Add last30days, agent-deep-research, hallmark, and impeccable to both skill trees. Pin vercel-labs/web-interface-guidelines so web-design-guidelines does not fetch mutable main. Extend the installer and catalog metadata. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Cursor indexes one-level SKILL.md directories. Link all packs (gstack first, awesome-copilot last), prefix collisions as <pack>-<name>, and leave OpenSpec-generated real skill dirs in place. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Initialize OpenSpec as the default spec/change layer, vendor Graphify, and add user-approved Context7 + Serena project MCP (no secrets). Document methodology XOR, last30days-as-leads, and Cloud Agent native-stack boot. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
…c trees Graphify full-tree index needs an LLM key because vendored SKILL.md files count as docs. Keep generated OpenSpec change/spec dirs and graph indexes out of git. Linker tests allow two minutes for a full flatten. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Keep Serena --context ide-assistant. Put bun on PATH for gstack setup, document Graphify pack-root SKILL.md and a non-interactive Serena index, and stop the linker test from requiring exact home vs project skill counts. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Current Serena lists ide, not ide-assistant. Keep Context7 + Serena as the only project MCP servers and document the name mapping. Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
Co-authored-by: Prax Lannister <praxstack@users.noreply.github.com>
|
Important Review skippedToo many files! This PR contains 1988 files, which is 1888 over the limit of 100. To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch. Upgrade to a paid plan to raise the limit. Usage-priced reviews support at most 300 files. ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1988)
You can disable this status message by setting the |
praxstack
marked this pull request as ready for review
August 30, 2026 16:28
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
📋 Description
Make Cursor on Cloud VMs pick up every vendored skill pack (not just gstack) as one-level
.cursor/skills/<name>/SKILL.md, and install the high-signal Aug 2026 PraxStack layers — OpenSpec, Graphify, Serena + Context7 MCP, last30days, agent-deep-research, Hallmark, Impeccable, pinned Vercel web-design-guidelines — without dumping 900 always-on personalities or specialist racks.Cursor indexes one level only.
scripts/link-agent-skills.shnow flattens all packs (gstack first, awesome-copilot last). Name collisions keep the first pack’s short slash name and prefix later packs (tdd→ pstack,mattpocock-tdd). OpenSpec-generated real skill dirs are never deleted.Methodology XOR: do not run gstack, Superpowers, pstack, and Compound Engineering as simultaneous always-on conductors. Flatten is discovery, not personality. OpenSpec is the default spec/change layer; Spec Kit stays vendored templates (no
specify initinto this Express app).gstack
./setup --host cursorfrom the slim vendor fails (scripts/resolve-codex-generation-model.tsmissing). Hostcursoris accepted (not gstack#2361). A full clone of gstack does complete setup. Durable Cursor path remains the flatten.Serena
--contextiside(current Serena built-in list; there is noide-assistant). Context7 has no API key in git.🔗 Related Issue
No issue. Follow-up to #38 (vendored packs), #40 (gstack flatten), #41 (Prax skills-and-personas, separate branch). Does not merge those PRs.
🔄 Type of Change
📸 Screenshots (if applicable)
Infrastructure only (Express review UI unchanged). Evidence:
cursor_native_setup_results.txt
npm_test_summary.txt
mcp_json_no_secrets.json
✅ Checklist
📝 Additional Notes
Installed: flatten of all existing packs; OpenSpec init; Graphify skill + Cursor rule; Serena + Context7 project MCP; last30days; 24601/agent-deep-research; Hallmark; Impeccable; pinned web-design-guidelines (
e3d624baaf29dc1fc645aff3e38f03e564d2d6b1); bun/uv/openspec/graphifyy on the VM.Bookmark only (not vendored): wshobson/agents, everything-claude-code, openai/plugins, NVIDIA/Remotion/Hyperframes/Manim/Stitch, Entire/Beads, other deep-research forks, remainder of microsoft/skills (~175).
Not done (documented): Graphify full LLM index (needs API key; would race flatten); Beads/Entire;
specify init..cursor/skillscount: 931. Required one-levelSKILL.md:plan-ceo-review,ce-brainstorm,improve,tdd,find-skills,last30days,hallmark,impeccable,graphify,openspec-propose,deep-research.npm test: 6 suites / 104 tests passed.npm run lint: 0 errors, 31 pre-existing warnings.Do not run two
link-agent-skills.shflattens (or Graphify + flatten) concurrently — the linker prunes then rebuilds.cursor/skills.To show artifacts inline, enable in settings.