adlc

Orchestration skill: chains the full ADLC Build Loop. PRD → Brief → Council → Scaffold → Codegen → LDD → TDD → Council → PR. Use when implementing a new feature end-to-end.

Generate a bounded, schema-valid goal prompt from one validated Build Brief task for zero-context external execution.

Applicability-aware Definition of Done checklist. Core checks apply to every task; overlay checks activate only when the applicability manifest says the surface exists. Binary verification — pass or fail. Triggers at Phase 4 completion and Phase 5 entry.

Score whether each task is specific enough for autonomous, one-shot, production-ready execution.

Reviews code changes for understanding, blast radius, state, secrets, assumptions, and explainability. Produces a blocking comprehension artifact when implications are unclear.

Resolve ambiguous triage outcomes in the middle confidence band.

Self-improving skill system. Captures diffs between agent output and human edits, distills patterns into rules, writes them back to skill files. Makes Build and Fix loops better over time.

Conditionally captures verified reusable ADLC learnings into docs/solutions after successful closeout.

Uses Graphify as ADLC's graph-backed research layer and Beads as an optional dependency-aware task memory layer. Produces evidence for compatibility, reuse, accuracy, dark-code hotspots, and long-horizon handoff.

Discovers and records repo-local approved build paths so agents reuse proven patterns instead of inventing parallel architectures.

Authors failing tests and reproducers from Build Brief acceptance criteria plus verification_spec before coding starts.

Assesses structural and velocity dark-code risk from architecture, AI tool usage, ownership, and deployment practices. Produces a direct risk assessment without inventing missing facts.

Orchestration skill for bug fixes and production issue repair. Investigate -> light brief -> targeted verification -> council -> PR.

Audits generated test strength on changed files using deterministic coverage and mutation measurement, then judges only surviving-mutant materiality.

Lint-Driven Development — formatting and syntax gate that runs BEFORE test execution. Violations block TDD entry. Triggers at Phase 4 entry.

Discovers existing functions/patterns/utilities that must be reused (not reimplemented) and identifies antipatterns to avoid. Triggers at Phase 1 (discovery) and Phase 5 (verification).

Scoped maintenance for docs/solutions entries when stale signals, refactors, or explicit user scope require refresh.

Builds structural, semantic, and philosophical context artifacts for a module or service: module manifest, behavioral contracts, and decision log.

Score generated prose or product output that cleared deterministic checks against a rubric threshold.

Orchestration skill for Magnus content delivery. Brief → Council → Draft → Slop Gate → Council(light) → Publish. End-to-end content creation and publishing.

Classify surviving mutants as trivial or material after deterministic mutation measurement.

Override deterministic section suppression when manifest evidence shows the section should stay active.

Decide whether an intersecting verifier actually exercises the semantic change.

Autonomous error detection and repair pipeline. Runs parallel to Build Loop. Capture → Confirm → Investigate → Fix → Prove → Light Council → Deliver or Escalate.

Logging mandate with applicability-aware activation: structured error, audit, and general logging are required only when the task introduces a runtime path, service boundary, or user-facing operation. Triggers at Phase 1 (specification) and Phase 5 (verification).

Security analysis with applicability-aware activation: STRIDE threat modeling only when the task introduces attack surface, trust-boundary, auth, data, or external-integration change; OWASP Top 10 vulnerability scanning per-diff post-execution (Phase 5).

Orchestration skill for Ratatosk trade execution. Thesis → Council → Risk Check → Execute → Audit → Report. End-to-end trade lifecycle.