[napetrov]

Stacked on #210 (feat/task-format-detect). Review the commits after #210: f7abf7c, 19822ba (precision in skill_selection_summary). The fix for vendored-skill discovery is a separate PR, #215.

Summary

dokimi could already write three trigger-set formats. With this PR it can read them too and run them through the existing skill-catalog: arm. It adds no new runner.

skill-reach QuerySet  ─┐
skill-creator / agentskills.io [{query, should_trigger}] =TARGET ─┼─> TriggerQuery ─> lint vs --catalog ─> questions | reach | skill-creator
questions + expected_skill ─┘                                                         └─> arms run --arms skill-catalog:<dir>  (skill_selection_summary)

 dokimi/external/trigger_set.py   + detect_trigger_format, from_reach, from_skill_creator, from_questions, read_trigger_set
 dokimi/commands/catalog.py       + trigger-set --trigger-set PATH[=TARGET];  + catalog formats --root [--json]
 dokimi/commands/arms.py          + a reach or skill-creator file gets the conversion command in its error
 dokimi/task_formats/             + consumer "catalog trigger-set" for skill-reach and skill-creator-trigger
 dokimi/skills/formats.py (new)   agent-skill (host + agentskills.io lint), cursor-rule, copilot-instructions, gemini-extension, claude-plugin

Evidence

• Before: arms run --questions reach.json failed in the questions loader, and tasks inspect gave no consumer for skill-reach or skill-creator sets.
  After: catalog trigger-set --trigger-set reach30.json --format questions reads the file and re-lints it (30 queries: 19 exam-framed, 17 echo). Arms then score it. The round-trip tests (writer, then reader, gives the same labels) are in tests/test_trigger_set_readers.py, and the package scan tests are in tests/test_skill_formats.py. The full suite passes locally (7834), along with ruff and mypy.
• skill_selection_summary now reports precision, which is hits over the questions where any skill loaded. The arms report shows it as a column. The precision figures below came from an ad-hoc script that used the same definition.
• Live run: haiku-4.5 (Bedrock) with the intel/skills@370d84d catalog (33 skills). Runs 1–3 are repeats.

condition	runs	recall	precision	false trigger
dokimi skill-catalog: with reach30, 33 skills	3×30	0.77	0.66	0.46
skill-reach claude-code haiku, same 30 ids (earlier run)	2×30	0.41–0.50	0.75	0.19–0.31
dokimi skill-catalog: with onetbb skill-creator set, 33 skills	1×20	1.00	0.56	0.80
dokimi skill-catalog: with onetbb skill-creator set, target alone	2×20	1.00	0.63	0.60
skill-creator run_eval.py native (anthropics/skills@dbd4588), haiku	3×20	0.10	n/a	0.00

The tool-calling proxy triggers too often compared with the real harnesses. The import path is unaffected by this; it is the same finding as before.

Merge Danger

Door: two-way

Additive only: two new CLI options or subcommands and a new module. --questions on catalog trigger-set is no longer required (at least one source is still enforced).

Blast Radius: small

A skill-creator negative is read as "no skill should load". In a catalog of more than one skill, loading a rival skill therefore counts as a false trigger. The CLI warns about this.

🤖 Generated with Claude Code

[github-actions]

CLI interface change detected

This PR changes dokimi's user-facing CLI surface (commands and/or options):

+ added command `dokimi catalog formats`
    + option `--help/-h`
    + option `--json`
    + option `--root`
+ added command `dokimi tasks inspect`
    + argument `path`
    + option `--help/-h`
    + option `--json`
~ `dokimi catalog trigger-set` parameters changed:
    + added option `--trigger-set`
    ~ changed option `--questions`: required: True -> False

Updated on every push; the cli-interface-change label is on while the diff is non-empty.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 90.81309% with 87 lines in your changes missing coverage. Please review.
✅ Project coverage is 91.97%. Comparing base (09be35b) to head (19822ba).

Files with missing lines	Patch %	Lines
dokimi/dokimi/external/trigger_set.py	81.91%	8 Missing and 9 partials ⚠️
dokimi/dokimi/task_formats/detect.py	87.61%	8 Missing and 5 partials ⚠️
dokimi/dokimi/task_formats/source.py	82.43%	8 Missing and 5 partials ⚠️
dokimi/dokimi/task_formats/datasets.py	94.54%	5 Missing and 7 partials ⚠️
dokimi/dokimi/skills/formats.py	92.90%	6 Missing and 4 partials ⚠️
dokimi/dokimi/task_formats/dirs.py	89.36%	4 Missing and 6 partials ⚠️
dokimi/dokimi/task_formats/configs.py	90.32%	3 Missing and 3 partials ⚠️
dokimi/dokimi/commands/catalog.py	95.38%	1 Missing and 2 partials ⚠️
dokimi/dokimi/commands/arms.py	83.33%	2 Missing ⚠️
dokimi/dokimi/commands/tasks.py	96.77%	0 Missing and 1 partial ⚠️
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #211      +/-   ##
==========================================
- Coverage   92.01%   91.97%   -0.05%     
==========================================
  Files         223      231       +8     
  Lines       27768    28701     +933     
  Branches     4563     4716     +153     
==========================================
+ Hits        25552    26398     +846     
- Misses       1570     1615      +45     
- Partials      646      688      +42     

Files with missing lines	Coverage Δ
dokimi/dokimi/__init__.py	100.00% <ø> (ø)
dokimi/dokimi/eval/arm_summary.py	97.41% <ø> (ø)
dokimi/dokimi/report/arms_report.py	96.95% <100.00%> (ø)
dokimi/dokimi/task_formats/__init__.py	100.00% <100.00%> (ø)
dokimi/dokimi/task_formats/model.py	100.00% <100.00%> (ø)
dokimi/dokimi/commands/tasks.py	92.40% <96.77%> (+1.06%)	⬆️
dokimi/dokimi/commands/arms.py	88.85% <83.33%> (-0.19%)	⬇️
dokimi/dokimi/commands/catalog.py	97.07% <95.38%> (-1.23%)	⬇️
dokimi/dokimi/task_formats/configs.py	90.32% <90.32%> (ø)
dokimi/dokimi/skills/formats.py	92.90% <92.90%> (ø)
... and 5 more

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.