adev-plugin

Assess codebase readiness for agentic development. Scans the codebase using static file inspection and outputs maturity scores across 8 structural dimensions (raw mode) or 11 dimensions (adev mode). Use when you want to evaluate how well-suited a codebase is for AI-assisted development, identify improvement areas, or track progress over time. In Codex, invoke with $adev:assess

Explore a feature idea interactively and produce a Feature Charter. Validates against constitution and existing charters. Use before building any new feature or module, when the user mentions a new capability, or says 'let us build X'. In Codex, invoke with $adev:brainstorm

End-to-end build orchestrator. Chains review, plan, route, implement, and validate for one or more specs through a full lifecycle pipeline. Use when the user says 'build', 'end to end', 'full pipeline', 'build the spec', 'build the milestone', 'run the whole pipeline', or wants to execute multiple lifecycle steps in sequence without manual handoffs. In Codex, invoke with $adev:build

Proactively scan source code for dead exports, orphan files, unused dependencies, stale code, and duplicate logic. Produces severity-tiered reports from repomap artifacts. Use when the user says 'check code health', 'find dead code', 'unused exports', 'stale files', 'orphan files', 'unused dependencies', or wants to identify cleanup opportunities before refactoring. In Codex, invoke with $adev:codehealth

Context-aware systematic debugging. Checks ADRs for known issues, specs for expected behavior, and orientation for architecture context before investigating. Use when encountering any bug, test failure, or unexpected behavior. Trigger on 'fix this', 'fix the bug', 'something is broken', 'this test fails', 'unexpected error', 'why is this not working', 'this doesn't work', 'can you fix', 'there's an error', 'it's failing', or any request to fix, debug, or investigate a problem. In Codex, invoke with $adev:debug

Run a deployment pipeline defined in .context-index/deploy.yaml. Executes shell, manual, verify, gate, and ci-trigger steps in order with failure recovery and rollback guidance. Use when the user says 'deploy', 'run deploy', 'push to production', 'publish', or wants to execute a structured deployment. In Codex, invoke with $adev:deploy

Generate human-readable developer documentation in docs/ from repomap output. Reads dependency-graph.json and symbol-ranks.json from /adev:repomap, plus charters and manifest. Use when the user says 'generate docs', 'document the codebase', 'create architecture docs', or wants to produce docs/architecture.md and per-module documentation. In Codex, invoke with $adev:document

Run a graduated evaluation harness (0-100) scoring implementation quality across four layers: deterministic, architectural, LLM-as-Judge, and human-in-the-loop. In Codex, invoke with $adev:eval

Audit all context for staleness, drift, and coverage gaps. Runs twenty audit passes across the .context-index/ directory and source code, generating actionable reports with checklists. Use when the user wants to check context health, find stale specs, detect drift between specs and code, identify missing coverage, scan for dead code, or clean up the context index. In Codex, invoke with $adev:hygiene

Execute implementation plans using specialist-routed subagents with TDD enforcement and 2-stage review per task. Use after planning to begin development. In Codex, invoke with $adev:implement

Initialize or diagnose the .context-index/ directory. Interactive wizard that explains each layer, detects existing setup, and lets users opt in or skip per layer. Use --brownfield for existing codebases, --dry-run to preview without writing. Trigger when the user wants to set up adev, initialize context, start a new project with adev, or diagnose a broken context-index. In Codex, invoke with $adev:init

Manage project issues and epics. Create, update, close, and view issues across file-based or beads_rust backends. Use when the user says 'create an issue', 'file a bug', 'show issues', 'what needs to be done', 'create epic', 'issue board', or wants to manage work items directly. In Codex, invoke with $adev:issues

Capture a lesson learned as a heuristic in the project's memory store. Use when the user says 'remember this', 'save this lesson', 'learn that...', 'file a heuristic', 'we learned that...', or shares a rule/pattern that should persist across sessions. In Codex, invoke with $adev:learn

Decompose reviewed Live Specs into ordered implementation tasks with TDD expectations and context routing. Use to break specs into actionable tasks. In Codex, invoke with $adev:plan

Rapidly sketch UI screens, user flows, and API surface from Feature Charters. Bridges the gap between chartering and implementation. Optionally uses a live browser preview for interactive design. In Codex, invoke with $adev:prototype

Interactive repair for lifecycle mismatches. Detects orphaned artifacts, stale epics, untraced code, and missing issues, then offers targeted fixes. Use when hygiene or status reveals inconsistencies, or when the user says 'reconcile', 'fix mismatches', 'clean up lifecycle', 'retroactive stamping', 'close stale epics'. In Codex, invoke with $adev:reconcile

Structured diagnosis-correction-resume cycle when agents get stuck during implementation. Classifies root causes into six categories, injects corrective context, and re-dispatches with enriched prompts. Writes recovery records for retrospective analysis. Use when a subagent is stuck, a task has stalled, an agent failed mid-execution, or the user reports 'the agent is looping' or 'it is not making progress'. In Codex, invoke with $adev:recover

Generate an AST-based symbol index of the repository. Extracts exported functions, classes, types, and interfaces, ranks by reference count, and outputs a repo map for drift detection by /adev:hygiene. Use when the user wants to map the codebase, generate a symbol index, understand the project structure, or prepare for drift detection. In Codex, invoke with $adev:repomap

Structured research skill. Investigates a topic using internal codebase search, web search, and GitHub code search, producing an organized research artifact. Use when the user says 'research', 'investigate', 'compare', 'how do others do this', 'best practices for', 'what are the options', or wants to explore approaches before committing to a design. In Codex, invoke with $adev:research

Analyze completed work over a time period to extract lessons, compute delivery metrics, identify improvement opportunities, and update context artifacts. Sprint retrospective for agentic development. Use when the user says 'run a retro', 'what went well', 'review the sprint', 'delivery metrics', or wants to reflect on recent development work. In Codex, invoke with $adev:retro

Run parallel specialist reviews (structural, security, consistency) on Live Specs before planning. Use for architecture review and expert evaluation. In Codex, invoke with $adev:review-specs

Score tasks on a four-dimensional routing matrix (spec completeness, pattern coverage, blast radius, novelty) and recommend auto-agent, assisted-agent, or human-only execution for each task in a plan. Use when the user asks 'which tasks need human review', 'score the tasks', 'route the plan', or wants to decide which tasks agents can handle autonomously versus which need human oversight. In Codex, invoke with $adev:route

Scan the codebase for high-quality implementations, score them against constitution and patterns, and curate annotated golden samples in .context-index/samples/. Golden samples serve as reference implementations that guide subagents during /adev:implement. Use when the user wants to find good examples, curate reference code, build a sample library, or identify high-quality patterns in the codebase. In Codex, invoke with $adev:sample

Author Live Specs within a Feature Charter's scope. Supports modes for new features, extraction from existing code, refactoring, diff-driven changes, and cross-cutting concerns. Use when the user says 'write a spec', 'define the behavior', 'create a contract', 'specify the feature', or needs to formalize requirements into a behavioral specification before planning. In Codex, invoke with $adev:specify

Query project status across charters, specs, capabilities, sessions, and source manifests. Read-only dashboard view of adev lifecycle artifacts. Use when the user asks 'what is the status', 'show project progress', 'which specs are done', 'charter status', 'capability progress', or wants a summary of where things stand. In Codex, invoke with $adev:status

Sync constitution to CLAUDE.md, AGENTS.md, and other agent files declared in manifest.yaml. Run after editing the constitution or when agent files are out of date. Use when the user says 'sync agent files', 'update agent files', 'constitution changed', 'regenerate agent configs', or after any edit to constitution.md. In Codex, invoke with $adev:sync

Gateway skill for the Agentic Development Framework. Injected at session start to establish methodology, available skills, and context routing. Use when the user asks 'what skills are available', 'how does adev work', 'what is the adev methodology', 'show me the workflow', or needs an overview of the framework and its lifecycle. In Codex, invoke with $adev:using-adev

Post-implementation validation with a trimmed code-time check set (quality gates, source manifest, code-drift advisory, spec compliance with scope-expansion sub-finding, constitution compliance with evidence-citation contract, optionally governance boundaries and transition gates, and visual verification for UI implementations). Fail-fast on quality gates. Structured PASS/FAIL report with migration-orientation footer pointing users to /adev:hygiene, /adev:reconcile, and /adev:review-specs for relocated concerns. Use when the user says 'validate the implementation', 'check if it works', 'run validation', 'verify the feature', or after implementation is complete and needs quality assurance. In Codex, invoke with $adev:validate

Pre-lifecycle triage that classifies incoming work and routes to the correct /adev:* skill. Scans for in-progress plans, unreviewed specs, and recent sessions before classifying. Use when the user says 'I need to work on something', 'what should I work on', 'what should I do', 'I have a bug', 'implement the feature', or any time the user is unsure which adev skill to invoke. In Codex, invoke with $adev:work

TDD test authoring: write failing tests (RED phase), produce immutable handoff blocks, and detect specification gaming. Use before implementation. In Codex, invoke with $adev:write-test

Assess codebase readiness for agentic development. Scans the codebase using static file inspection and outputs maturity scores across 8 structural dimensions (raw mode) or 11 dimensions (adev mode). Use when you want to evaluate how well-suited a codebase is for AI-assisted development, identify improvement areas, or track progress over time. In OpenCode, invoke with skill({ name: 'adev:assess' })

Explore a feature idea interactively and produce a Feature Charter. Validates against constitution and existing charters. Use before building any new feature or module, when the user mentions a new capability, or says 'let us build X'. In OpenCode, invoke with skill({ name: 'adev:brainstorm' })

End-to-end build orchestrator. Chains review, plan, route, implement, and validate for one or more specs through a full lifecycle pipeline. Use when the user says 'build', 'end to end', 'full pipeline', 'build the spec', 'build the milestone', 'run the whole pipeline', or wants to execute multiple lifecycle steps in sequence without manual handoffs. In OpenCode, invoke with skill({ name: 'adev:build' })

Proactively scan source code for dead exports, orphan files, unused dependencies, stale code, and duplicate logic. Produces severity-tiered reports from repomap artifacts. Use when the user says 'check code health', 'find dead code', 'unused exports', 'stale files', 'orphan files', 'unused dependencies', or wants to identify cleanup opportunities before refactoring. In OpenCode, invoke with skill({ name: 'adev:codehealth' })

Context-aware systematic debugging. Checks ADRs for known issues, specs for expected behavior, and orientation for architecture context before investigating. Use when encountering any bug, test failure, or unexpected behavior. Trigger on 'fix this', 'fix the bug', 'something is broken', 'this test fails', 'unexpected error', 'why is this not working', 'this doesn't work', 'can you fix', 'there's an error', 'it's failing', or any request to fix, debug, or investigate a problem. In OpenCode, invoke with skill({ name: 'adev:debug' })

Run a deployment pipeline defined in .context-index/deploy.yaml. Executes shell, manual, verify, gate, and ci-trigger steps in order with failure recovery and rollback guidance. Use when the user says 'deploy', 'run deploy', 'push to production', 'publish', or wants to execute a structured deployment. In OpenCode, invoke with skill({ name: 'adev:deploy' })

Generate human-readable developer documentation in docs/ from repomap output. Reads dependency-graph.json and symbol-ranks.json from /adev:repomap, plus charters and manifest. Use when the user says 'generate docs', 'document the codebase', 'create architecture docs', or wants to produce docs/architecture.md and per-module documentation. In OpenCode, invoke with skill({ name: 'adev:document' })

Run a graduated evaluation harness (0-100) scoring implementation quality across four layers: deterministic, architectural, LLM-as-Judge, and human-in-the-loop. In OpenCode, invoke with skill({ name: 'adev:eval' })

Audit all context for staleness, drift, and coverage gaps. Runs twenty audit passes across the .context-index/ directory and source code, generating actionable reports with checklists. Use when the user wants to check context health, find stale specs, detect drift between specs and code, identify missing coverage, scan for dead code, or clean up the context index. In OpenCode, invoke with skill({ name: 'adev:hygiene' })

Execute implementation plans using specialist-routed subagents with TDD enforcement and 2-stage review per task. Use after planning to begin development. In OpenCode, invoke with skill({ name: 'adev:implement' })