Command
Part of truestack
Run a verified, multi-source research pass and return a cited answer — parallel searches, primary sources, adversarial cross-checking, confidence stated.
Honesty-first coding skill set for Claude Code — 21 skills behind an orchestrate router with an enforced governance gate, a code↔memory tally, and auto-research. Built for self-hosted work.
Command
Part of truestack
Run a verified, multi-source research pass and return a cited answer — parallel searches, primary sources, adversarial cross-checking, confidence stated.
Command
Part of truestack
Score and audit Agent Skills — static lint + semantic judge + behavioral trigger test — and report a scorecard with concrete fixes.
Command
Part of truestack
Iterate to the goal — keep building/fixing and re-running truestack-quality-control until the acceptance criteria pass (or report an honest completion score if blocked).
Command
Part of truestack
Study this repo and build (or refresh) its committed memory — stack, commands, conventions, boundaries, and the always-on contracts — so every other skill reads it first. Replaces a manual onboarding pass; named onboard to avoid colliding with the built-in /init.
Command
Part of truestack
Run the project's tests / build / lint commands for the detected stack and report results in a scannable table.
Command
Part of truestack
Set up a recurring or deferred automatic run — define the trigger, the self-contained run prompt, delivery, and failure policy, then wire it to the host scheduler.
Command
Part of truestack
Run a maintainability simplify pass on the target — merge duplication, drop needless layers and dead code, clarify names — without changing behavior.
Command
Part of truestack
Run a full quality sweep (truestack-quality-control) on the current changes — tests, types/lint, 6-axis review, load/perf, safety, and intent — and return a verdict.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: