en un clic
dss-code-prime
dss-code-prime contient 5 skills collectées depuis dailysoftwaresystems, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Advance the DSS Code Prime compiler by exactly ONE development cycle: pick the next priority from the plan-00 §0.1 stepper, clear its blockers first, plan it with /feature-dev:feature-dev, design-audit the plan against the bar before locking it, implement the best long-term solution, run /pr-review-toolkit:review-pr (and re-review the fold if it changed logic), pin every deferral with a priority, cross-plan update, self-audit the cycle against the bar before locking, then commit and push. One cycle per invocation — meant to be driven by /loop for continuous autonomous progress. PAUSES and asks the user on any pending definition, architectural fork, gated anchor, or hard-stop boundary; never guesses, never workarounds, never breaks source (language) / target (processor) / linker (object-format) agnosticism.
Measure the current state of the DSS Code Prime compiler. Runs a 104-probe C-feature battery through the real dss-code-prime CLI and reports every % axis — empirical C coverage, plan-23 full-C arc %, SQLite-readiness %, test suite, deferred-anchor closure, cross-target emit/run matrix — plus a velocity-based ETA for "compiles SQLite" and the full C23 arc. Use when asked how far along the compiler is, for a progress/state/status report, % complete, an ETA for full C or SQLite, or to re-measure after a /dss-cycle.
DSS Code Prime repository guide — universal config-driven C++ compiler frontend. Covers the tree/node model, schema-driven grammar configs, TreeBuilder, TreeCursor + visitor walks, NodeAttribute<T> side-tables, typed views, diagnostic discipline, the .plans/ system, CMake build wireup, coding conventions, and (critically) the strict-assertion testing posture that makes every regression visible.
Audit the DSS Code Prime compiler against the bar — the read-only manager/auditor counterpart to dss-cycle. Independently verifies the implementer's output (status checks, green/registry/agnosticism battery, guardrail enforcement, silent-gap hunting) and renders a verdict that separates VERIFIED-CLEAN from GREEN-BUT-RULE-BREAKING. Also design-reviews a developed plan against the bar before it is locked for build (dss-cycle Step 3.5). Never builds, never edits src/ — its integrity comes from not having written the code it judges. May commit ONLY markdown, separately tagged — trivial plan/doc staleness, and test-discipline ratchets into the dss-cycle skill on a recurring weak-test class. On any gap it authors the exact ready-to-paste implementer prompt. Judges by: best long-term / no workarounds / source (language) + target (processor) + linker (object-format) agnostic / fail-loud — and holds ITSELF to "verified, not attested": green is never clean until independently re-run.
Comprehensive plan-staleness reconciliation for DSS Code Prime — sweep EVERY plan in .plans/ (+ the deferred-anchor registry, README, sibling skills) and reconcile every status, count, description, anchor, and cross-reference against the CURRENT observable implementation (git + ctest + src/ + the anchor guard). The third sibling: dss-cycle IMPLEMENTS, dss-audit JUDGES code, dss-plan-sweep RECONCILES plan-truth. It FIXES — markdown only, never src/ — everything reconcilable to observable fact (drifted counts, already-pushed "pending" markers, duplicate/contradictory rows, broken refs, numbering drift, a status that contradicts another plan surface). For doneness JUDGMENTS (striking an anchor closed, rewriting a description to match new code) it fixes only on unambiguous evidence, else FLAGS. Its guarantee: it leaves NO staleness behind — every divergence is fixed-with-evidence or explicitly flagged, never silently ignored, never over-claimed. Reconciles to truth, never to opinion; never deletes a historical/lo