pe-menezes/vibeflow

Deep-analyzes the current codebase to discover stack, architecture, patterns, conventions, and pitfalls. Creates curated documentation in .vibeflow/ that persists and can be committed to git. Supports incremental updates, scoped deep-dives, interactive review, and satellite repo analysis. Use when setting up a project, after significant changes, or to deep-dive into a specific module.

Audits recent work against its Definition of Done and project patterns, from a review package of the diff. Runs the test suite, compares code against the spec, tests scenarios, sabotages risk code, and reports PASS / PARTIAL / FAIL with a severity and a cause on every finding. Also runs the Critical Gate — a safety scan of the diff for destructive or dangerous operations. Generates an incremental prompt pack for any gaps found. With --reaudit, checks a fix round against its own diff; with --consolidate-hotfixes, reclassifies hotfix trace docs against the current code. Use after implementation to verify quality before shipping.

Runs an interactive discovery dialogue (1–5 rounds) to turn a vague idea into a clear, actionable PRD. Challenges assumptions, cuts scope, and forces decisions. Saves the PRD to .vibeflow/prds/. Use before gen-spec when the idea is not yet well-defined or requirements are unclear.

Generates a technical spec with objective, Definition of Done (3–7 binary checks), scope, anti-scope, technical decisions, and applicable patterns. Grounded in the project's real patterns from .vibeflow/. Auto-splits specs that exceed budget or DoD limits. Use when requirements are clear and you're ready to define the implementation contract.

Fixes an observed defect with reproducible evidence in one call: writes a short trace doc before touching code, implements the fix, and backs it with a regression test written before the fix. Production incidents are the motivating case, not a gate. When blocked, it halts by name and saves the doc for a later call to resume. Use when a defect has concrete evidence and you need the fix with its paper trail now — planned small tasks belong to quick, spec'd features to implement.

Implements a feature from its spec following all guardrails: budget, DoD, anti-scope, and pattern compliance. Runs a 9-phase pipeline, phases 0 to 8 (find spec → extract guardrails → load patterns → plan → implement → test → refine → self-verify DoD → finish). Use after gen-spec when you're ready to code. The agent has filesystem access and acts as Coding Agent.

Generates a self-contained prompt pack from a spec. Embeds real code patterns from .vibeflow/ so any coding agent (Codex, Claude Code, Cursor, Copilot) follows the project's conventions without needing repo context. Use when handing off implementation to a separate session or agent.

Fast-tracks small tasks into a ready-to-use prompt pack in one command. Skips discover, generates an ephemeral spec in memory, and outputs a prompt pack with embedded patterns. Use for well-defined tasks that fit in ≤4 files.

Compiles statistics from audit reports: PASS/PARTIAL/FAIL rates, most violated patterns, most failing DoD checks, and quality trends over time. Output is chat-only (no file created). Use after running several audits to spot quality patterns and improvement areas.

Updates the project knowledge base (.vibeflow/) with corrections, new conventions, architectural decisions, or new patterns based on natural language feedback. Also imports patterns from external repos through the --from option. Edits are placed outside auto-generated markers to survive incremental analyze runs. Use to keep .vibeflow/ accurate as the project evolves.