Menu
Skills in this repository
SkillsMP has collected 178 skills from DylanCkawalec/opgrok. Open a skill to review its source and details.
DylanCkawalec/opgrokShowing 40 of 178 collected skills.
Audits SuperGrok skill catalogs, registries, and program hygiene via explicit checklist scoring with path:line evidence. Activates on regenerate-after-role-change, rebind-identities, registry repair, or /meta-audit. Differentiator: enforces the…
Authors and hardens SuperGrok skills, registry, and catalog hygiene via the forge method: full generate→rebuild→identity→validate path before edge fixes. Activates on catalog drift, role/SKILL.md bulk edits, registry rebinds, or /meta-forge. Differentiator:…
Maps SuperGrok catalog topology, registry drift, and skill-authoring constraints before any edit lands. Activates on regenerate-after-role-change, identity rebind, registry repair, or /meta-scout. Differentiator: pre-commit structure map that treats…
Finalizes SuperGrok catalog and registry work: runs the Leslie win gate, freezes generator outputs, and marks artifacts handoff-ready. Use after role edits, category adds, bulk skill changes, or /meta-seal. Differentiator: enforce…
Routes work to the correct SuperGrok role inside the meta domain (skill authoring, registry, asset systems). Activates when selecting a meta specialist, browsing meta agents, or invoking /cat-meta. Differentiator: category index only — does not perform meta…
Authors and repairs SuperGrok skills, registry entries, and catalog hygiene as atomic build units. Activates on regenerate-after-role-change, identity rebind, registry drift, or /meta-smith. Differentiator: enforces the generate→rebuild→ identity→validate…
Root-causes SuperGrok catalog, registry, and skill-hygiene failures via symptom→evidence→root→fix chains. Activates on broken validate gates, stale identities after bulk SKILL.md edits, registry drift post-category add, or /meta-trace. Differentiator: forces…
Audits design docs, ADRs, PR plans, and delivery sequences against an explicit quality checklist, scoring pass/fail per item with path:line evidence. Use for storage-choice ADRs, multi-package PR plans, migration sequences with rollback gates, or when invoked…
Forges implementable design docs, ADRs, PR plans, and delivery sequences by locking the full end-to-end path before edge hardening. Use for storage ADRs with options+decision, multi-package PR plans, or migration sequences with rollback gates; activates on…
Maps repo structure, constraints, and decision surfaces before any plan or ADR is written. Activates for design docs, ADRs, PR plans, delivery sequences, or when invoked as /plan-scout. Differentiator: refuses pathless vibe diagrams; every map names…
Finalizes design docs, ADRs, PR plans, and delivery sequences: verifies win-gate evidence, freezes file-concrete outputs with owners/risks/acceptance, marks handoff- ready. Activates for ADR storage decisions, multi-package PR plans, migration sequences with…
Routes work to the correct SuperGrok role inside the plan domain (design docs, ADRs, PR plans). Activates when selecting a plan specialist, browsing plan agents, or invoking /cat-plan. Differentiator: category index only — does not perform plan implementation…
Drafts file-concrete design docs, ADRs, PR plans, and delivery sequences as the smallest correct build unit that meets the brief. Activates on ADR/storage-option decisions, multi-package PR plans, migration sequences with rollback gates, or /plan-smith.…
Builds implementable design docs, ADRs, PR plans, and delivery sequences by tracing symptom → evidence → root → fix. Activates for storage/ADR decisions with options, multi-package PR plans, migration sequences with rollback gates, or /plan-trace.…
Audits product specs for hireable clarity: must/should/could, explicit non-goals, and observable PASS/FAIL acceptance criteria. Use when turning vague asks into one-pagers, backlog ranks, or harness-ready AC, or when invoked as /product-audit. Differentiator:…
Turns vague product asks into hireable specs: problem, users, MoSCoW + non-goals, observable acceptance, and a phased e2e path before edge hardening. Use for one-pagers, backlog risk/value cuts, or harness-ready acceptance criteria. Triggers on /product-forge…
Maps product structure, constraints, and ambiguity before any spec edit or backlog commit. Use when turning vague asks into MoSCoW + non-goals, hireable acceptance criteria, or risk/value prioritization. Triggers on /product-scout and phrases like one-pager,…
Finalizes product specs into hire-ready artifacts: MoSCoW cuts, explicit non-goals, and observable PASS/FAIL acceptance. Use when freezing a one-pager, turning a vague ask into harness criteria, or ranking backlog by risk×value before plan/code handoff.…
Routes work to the correct SuperGrok role inside the product domain (specs, prioritization, requirements). Activates when selecting a product specialist, browsing product agents, or invoking /cat-product. Differentiator: category index only — does not perform…
Turns vague product asks into hireable build units: one-pager with MoSCoW, explicit non-goals, and observable PASS/FAIL acceptance. Activates for Write a one-pager with must/should/could and non-goals, backlog risk/value cuts, or /product-smith.…
Traces product failures and vague asks into hireable specs via RCA: symptom → evidence → root → fix. Use for one-pagers with must/should/could + non-goals, backlog risk/value cuts, or turning fuzzy requests into observable acceptance criteria. Activates on…
Audits Python packages, typed APIs, and async services against a scored checklist (types, awaits, excepts, layout, tests, secrets). Activates on /python-audit or tasks like typed FastAPI DI routes, un-awaited coroutine races, and pytest extraction of pure…
Ships Python packages, typed APIs, and async services by forging the full call path (entry → service → data) before edge hardening. Activates on /python-forge and tasks like typed FastAPI DI routes, un-awaited coroutine races, or extracting pure helpers under…
Maps Python package topology, typing surface, async boundaries, and test layout before any edit. Activates on /python-scout and tasks like typed FastAPI DI routes, un-awaited coroutine races, or pytest extraction of pure helpers. Differentiator:…
Finalizes Python package and service changes: verifies win-gate evidence, freezes typed public surfaces, and marks artifacts ready for handoff. Activates on /python-seal or when closing FastAPI/async/pytest work. Differentiator: import-graph-aware seal that…
Routes work to the correct SuperGrok role inside the python domain (Python packages, typing, async). Activates when selecting a python specialist, browsing python agents, or invoking /cat-python. Differentiator: category index only — does not perform python…
Builds the smallest correct Python unit—typed package surface, async service path, or script—against existing layout and deps. Activates on /python-smith or briefs like typed FastAPI DI routes, un-awaited coroutine races, pytest-covered pure helpers.…
Root-cause traces failing Python packages, typed APIs, and async services by chaining symptom → evidence → root → minimal fix. Use for traceback-driven repair, un-awaited coroutines, import/layout breaks, or /python-trace. Differentiator: pytest node-id +…
Runs multi-source research and grounded synthesis under an explicit audit checklist, scoring each criterion PASS/FAIL with cited evidence. Use for competitive landscapes with dated sources, codebase “where is X handled?” traces, or contradiction-aware…
Multi-source research and grounded synthesis for competitive landscapes, codebase locus questions, and contradiction-aware briefs. Builds the full e2e evidence path before edge hardening; every hard claim carries URL or repo path, with explicit…
Maps multi-source research before synthesis: web/code search, dated citations, fact/inference split, contradiction callouts. Activates for competitive landscape with dated sources, "where is X handled?" codebase hunts, or /research-scout. Differentiator:…
Finalizes multi-source research packets: audits citation coverage, freezes the fact/inference/opinion split, and emits a handoff-ready brief with dated sources. Activates for competitive landscapes, codebase "where is X" answers, contradiction synthesis, or…
Routes work to the correct SuperGrok role inside the research domain (multi-source grounded research). Activates when selecting a research specialist, browsing research agents, or invoking /cat-research. Differentiator: category index only — does not perform…
Builds the smallest cited research unit that answers one brief: multi-source web/code retrieval, fact/inference split, contradictions surfaced with dates. Activates on competitive landscapes, "where is X handled?", synthesis with source conflict, or…
Builds cited causal chains (symptom → evidence → root → fix) across web and codebase sources for competitive landscapes, "where is X handled?", and multi-source synthesis with contradictions surfaced. Activates on /research-trace or dated multi-source…
Audits code/design changes against an explicit PASS/FAIL checklist, emitting severity-ordered findings with path:line evidence. Use for PR correctness and test-gap reviews, ADR risk ranking, or API blast-radius critique; triggers on /review-audit.…
Severity-ordered code and design review via the forge method: reconstruct the full end-to-end execution path before edge hardening. Activates on PR correctness/test-gap reviews, ADR risk ranking, API blast-radius critiques, or /review-forge. Differentiator:…
Maps PR/design structure and constraints before ranking findings for code and design review. Activates on PR correctness/test-gap reviews, ADR risk ranking, API blast-radius critiques, or /review-scout. Differentiator: builds a touch-map (entrypoints,…
Seals code/design reviews by severity-ranking findings with path:line evidence, verifying win-gate checks, freezing the report, and marking handoff-ready. Activates on PR correctness audits, ADR risk ranking, API blast-radius critique, or /review-seal.…
Routes work to the correct SuperGrok role inside the review domain (code and design review). Activates when selecting a review specialist, browsing review agents, or invoking /cat-review. Differentiator: category index only — does not perform review…