[SakarPaisen]

What

Closes #14.

convert_timestamp now returns weekday_utc, the English weekday name of the converted UTC date, in both directions.

>>> convert_timestamp("0")
{'unix_seconds': 0.0, 'iso_utc': '1970-01-01T00:00:00Z', 'weekday_utc': 'Thursday'}
>>> convert_timestamp("1970-01-02T00:30:00+01:00", "to_unix")
{'unix_seconds': 84600.0, 'iso_utc': '1970-01-01T23:30:00Z', 'weekday_utc': 'Thursday'}

Scope

• src/mcp_tools_lab/tools/timestamps.py — a fixed WEEKDAYS tuple indexed by date.weekday(), plus a docstring line.
• tests/test_timestamps.py — four new tests.
• docs/tool-guide.md — documents the field and the midnight-crossing example.

The name comes from a fixed list rather than locale formatting, so it does not depend on the machine's language settings. Existing values and error handling are unchanged.

Tests added

• Known weekdays: epoch "0" → Thursday; "345600" → Monday; "1970-01-04T00:00:00Z" in to_unix → Sunday.
• Offset crossing midnight: "1970-01-02T00:30:00+01:00" is 1970-01-01T23:30:00Z → Thursday, even though the input's local date is Friday.
• Negative timestamp: "-0.5" → Wednesday.
• Both directions agree for the same instant.

Millisecond mode

#6 (millisecond support) has not landed: convert_timestamp still takes only value and direction. There is therefore no second output mode to cover yet; when #6 lands, weekday_utc is already part of the single return statement, so it will apply there as well.

Checks

• python -m pytest tests/test_timestamps.py -q — 13 passed
• python -m pytest -q — 66 passed
• python examples/try_tools.py — runs and shows 'weekday_utc': 'Thursday'
• python .github/scripts/check_rules.py — Scanned 11 files under src/ for print() and network imports: OK
• Direct calls for "0", the midnight-crossing offset, and "-0.5", printed and compared with the issue's examples

Python: 3.13 on Windows 11.

(Full disclosure: I used an AI coding assistant for this change, and I reviewed and verified everything submitted here.)