Artifact-first agent framework — disciplined skill chain for AI-driven multi-step engineering tasks. MIT.
Skill Claude CodeCodex
Session startup skill. Reads platform docs, builds context, asks which project we're working on, reads project docs — and is ready to work. Also handles continuation of a stuck task by restoring context from the interrupted session. Use when: start of a new terminal session, load context, resume interrupted work.
Skill Claude CodeCodex
Protocol for architecturally clean execution of complex multi-block tasks. Decomposition into blocks → research before library-first → explicit decision pinning → clear scope vs out-of-scope → report after each block. Principle: every architectural decision is pinned explicitly, no ambiguity, epics don't sprawl. Use…
Skill Claude CodeCodex
Automatic linking of an artifact to project architecture elements. Two modes: file-mode (by file paths, for fast-track flows) and feature-mode (by artifact text, for briefs/specs). Merge logic: appends to existing archref, does not overwrite. Propagate: final archref written up the chain epic → spec → brief. Use when…
Skill Claude CodeCodex
Audit-before-fix protocol — first find all gaps across 7 planes, prioritize, pin the gap table, and only then fix. Principle: never touch code until the full picture of problems is understood. Gap table → approval → flow-first per gap → library-first → plan → code. Use when: task begins with looking for problems…
Skill Claude CodeCodex
Meta-orchestrator: assigns skill-chain cadence per block (Tier 1/2/3) via Q1-Q6 rules. Reads target taskblocks + task.md, writes cadence-decisions-{R}.md artifact. Standalone (executor invokes) or Batch (generator hand-off). Use when: deciding cadence for a batch of blocks, generator hand-off from decomposition skills.
Skill Claude CodeCodex
Coverage validation protocol before a final approval gate. Reads requirements from any structured requirements document, matches them against the proposed decomposition units (blocks, epics, tasks), outputs a coverage table. If gaps exist — stops, waits for user decision. If everything is covered — continues without…
Skill Claude CodeCodex
Makes an architectural / project / scope decision using a 5-part model INSTEAD of asking the user. Structure: 🎯 Decision / Why / 🛡 Security / 📈 Scalability / Alternatives / Plain-language analogy. 1 question = 1 atomic artifact. Use when: the agent is about to ask an architectural / scope question…
Skill Claude CodeCodex
Autonomous orchestrator of the per-block go-fast cycle. Takes over after arch-first/audit-first when "autopilot" mode is chosen. Runs every block: flow-first → library-first → plan-first → execute without manual approvals. Validates each skill's artifact against built-in rules. On blockers — escalates to a chat bot…
Skill Claude CodeCodex
Understanding-alignment protocol before library-first. Asks the user for 2-3 anchors (file, table, route, service), reads only those (not the whole project), fills a 4×3 table (Landscape / Problem / Solution / Result × UI / DB / Integrations), waits for approval — and only then hands off to library-first. Use when…
Skill Claude CodeCodex
Adds a new conceptual guide to a project's knowledge folder. Runs in three modes: interactive (user-invoked), from-task (auto-invoked from ship-first after a fast-track completes), from-epic (from epic closure). Storage: local .md file under project-knowledge/guides/{topic}/, registered via git commit + push. No…
Skill Claude CodeCodex
Takes the agent's last message and explains it in plain language — without technical jargon, with concrete examples and analogies. Invoked manually when the user did not understand what the agent just wrote. Use when: "/human-first", "explain in plain language", "didn't get what you wrote", "explain simpler".
Skill Claude CodeCodex
Entry point of the self-organizing skill chain. Launched right after task creation. Asks 5-7 questions, determines task type (product / feature / fix), waits for approval — then proposes the next skill per branch: product → habit-first, feature → arch-first, fix → audit-first. Principle: without a defined type, the…
Skill Claude CodeCodex
Mandatory protocol before executing any fast-track task. Analyzes the task, builds a table: what we do / where it comes from / how many lines of code. Principle: maximum reuse of existing libraries and components, minimum new code. Waits for explicit user approval — does nothing until confirmed. Use when: fast-track…
Skill Claude CodeCodex
Saves the last assistant message as a note attached to the current task. Auto-numbered (note-01, note-02...), bidirectional link with task.md. Invoked manually by the user. Use when: "/note-first", "save a note", "remember this", "save the last message".
Skill Claude CodeCodex
Mandatory protocol before any document creation, code writing, refactoring, DB migration, test plan, or git operation. Shows a plan table → picks a mode → waits for confirmation → executes step by step. Use when: before writing/refactoring code, migrating DB, writing tests, git commit/PR; "/plan-first", "make a plan…
Skill Claude CodeCodex
Final task completion protocol: report → user-note → deploy → smoke test → close? → guide? → routing.db → propagate → STATUS → sessions. Invoked from fast-track and pipeline flows after task execution, not directly by the user. Use when: invoked after the execute step of a fast-track or pipeline task.
Skill Claude CodeCodex
Homework validator — checks that user completed the QUICKSTART tutorial correctly. Runs 5 SQL/file checks (routing.db populated, project registered, demo workspace, skill invocations logged, task folder structure) plus an optional 6th self-audit step. Reports ✅/❌ per check with remediation hints. Use when: after…
Skill Claude CodeCodex
Final audit of a large task before closure — finds which operations are available only via code / curl / SQL and decides per each: automate with a skill or AI agent (A) or create a UI task (B). Protects against invisible usability debt. Walks through each implemented block: reads task.md + reports + guides + code →…
Skill Claude CodeCodex
Creates a parity fixture — the frozen scenario plus the contract its output must satisfy. Asks which skill and case, what shape the run must produce, and writes input.md and expect.yml. Ends by proving the new fixture actually fails on an empty directory. A fixture that passes when nothing ran is worse than no…
Skill Claude CodeCodex
Thin entry point into a self-organizing skill chain for a fast-track task. Creates a task (slug, number, task.md, routing.db entry) and passes control to idea-first. The chain then runs autonomously via Skill() calls. Use when: "/go-fast", start a fast-track task, quick task with autonomous routing through the skill…
Skill Claude CodeCodex
Guides a fixture run end to end and reports whether the framework behaved the same on this runtime as on the reference one. Picks a fixture, hands its task statement to the agent under test, waits, locates the log directory the run produced, then calls bin/parity-run and explains the result. The stop in the middle is…
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: