Loading...
Loading...
Found 8 Skills
Audit test coverage for code changes. Identifies untested logic and provides specific test recommendations. Read-only analysis. Use before PR or after implementation. Triggers: review coverage, check tests, test coverage, are tests adequate.
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
Explain an existing Convex app — data model + relationships, public vs internal functions, auth/ownership model, components, a request→data flow — read from the schema and function surface. Read-only.
Check and validate MTHDS bundles for issues. Use when user says "validate this", "check my workflow", "check my method", "does this .mthds make sense?", "review this pipeline", "any issues?", "is this correct?". Reports problems without modifying files. Read-only analysis.
Check and validate MTHDS bundles for issues. Use when user says "validate this", "check my workflow", "check my method", "does this .mthds make sense?", "review this pipeline", "any issues?", "is this correct?". Reports problems without modifying files. Read-only analysis.
Trace and explain how an existing code path or subsystem works.
Produce a ranked contribution table from a batch of analyzed projects. Read a parent Multica issue whose sub-issues (scouts) each analyzed one project and posted a deliverable comment with top picks, extract every top-pick contribution, score each against the mmontes11/contributions #contribution-scoring rubric, rank descending, and post the table on the parent issue. Use whenever the user wants a cross-project ranking of the contributions the scouts surfaced — "rank the contributions", "build the contribution scoring table", "which contributions are most likely to be merged first", "produce the contribution ranking for the analyzed projects". Read-only against the repos — it never forks, edits code, or opens PRs.
GPT (GPT-6 Astra by default) as advisor, adversarial reviewer or judge through headless `codex exec`, always read-only; never an implementer. Modes: advise (independent read on a plan, diff or design, beside Claude's own view, with follow-ups), review (Codex's built-in reviewer on uncommitted work, a branch or a commit), adversarial (tries to break confidence in a change; optional parallel lenses; every finding verified by Claude before it is shown), judge (scores a reviewer's verdict for insistir and goal-loop). Trigger on "second opinion", "segunda opinión", "ask codex/GPT to review", "que codex lo revise", "revisión adversarial", "adversarial review", "que GPT ataque este cambio", "critica este plan", "challenge this design", "judge this review", "usa astra". Falls back to a Claude subagent, labelled same-provider, when Codex is unavailable. Not askcodex (one-shot text or images) and not for writing code.