lazyai

Use when debugging a failing test, broken build, runtime error, or unexpected system behavior. Drives hypothesis-based investigation, root-cause fixes, verification, and reusable learning capture.

Use when a claim, plan, spec, doc, or design must be attacked against source evidence before implementation or approval. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when fixing broken behavior and the priority is reproduce, root cause, regression proof, and the smallest safe repair. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when reviewing code changes with explicit stage, target, and focus so own-code checks and others’ PR reviews stay distinct but share one contract. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when the deliverable is docs or reference material and every claim must stay tied to observed behavior or cited source evidence. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when adding or changing product behavior that must ship with explicit purpose, scenarios, and verification. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when improving structure without changing behavior, with explicit invariants and verification that the contract stays the same. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when the goal is to investigate, map constraints, and return evidence without committing to implementation. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Disciplined investigation methodology with verification gates and append-only contribution back. Nine-phase pipeline that plans, parallelizes, verifies citations, cross-references authoritative team docs, re-baselines when needed, runs adversarial review, and contributes findings without overwriting others' work. Uses write tools, shell execution, and may spawn subagents. Invoke with enable_write_tools=true and enable_subagent_tools=true.

Use when you need a Slack-ready rewrite of a longer message. Produce sharp, legible, correctly spelled Slack mrkdwn and keep it concise enough to skim.

Use when turning markdown-heavy or code-heavy text into Slack-optimized formatting. Keep text short, preserve intent, and use npx @slackfmt/cli without making the project depend on npm.

Use when entering an unfamiliar repository or subsystem and you need a disciplined search-and-read strategy before changing code.

Use when asked to create or revise a LazyAI agent source definition while keeping generated tool-native agent files derived, not hand-edited.

Use when asked to create or revise a LazyAI hook policy so runtime enforcement stays objective, generated, and tool-specific.

Use when asked to create or revise a LazyAI skill source file while keeping generated tool-native skill surfaces derived, not hand-edited.

Use when asked to create or revise a LazyAI workflow artifact as documentation/catalog guidance without reintroducing a runtime workflow engine.

Use during implementation to run the smallest meaningful verification command after each focused change.

Use when starting or steering an agent task with four-point communication, project constitution, and fast feedback.

Use when creating, renaming, or modifying LazyAI skill source files so the canonical source and generated tool-native outputs stay aligned.

Verify plan completeness via architectural & requirement traceability analysis.

Gate implementation against spec and constitution. Verify all requirements met and all gates passed.

Resolve ambiguity in spec.md through sequential, focused questioning. Records answers in the spec's Clarifications section.

Establish or amend the project constitution — the governing contract every workflow is judged against.

Execute a task following its harness contract and 5-gate loop.

Author an implementation plan from a validated spec. Defines approach, phases, and measurable milestones.

Author a feature specification from a user description. Output is the contract every downstream artifact is judged against.

Break a plan into executable task files with dependency graph and parallelization.

Use when extracting actionable tasks from meeting notes, Slack threads, PR descriptions, specs, or other unstructured text. Converts ephemeral notes into tracked issues with context, acceptance criteria, deduplication, and learning capture.

Index the codebase with codegraph + qmd, build a knowledge graph with graphify, and fill remaining <!-- fill-in: --> placeholders in AGENTS.md with detected architecture, conventions, patterns, and rules. Every filled value is tagged with confidence (🟢🟡🔴).

Use when asked to create or revise a LazyAI agent source definition while keeping generated tool-native agent files derived, not hand-edited.

Use when asked to create, scaffold, or write a new vibe-lab hook or hook policy. Defines one canonical POLICY.md and optional runtime scripts, with Claude Code hook and OpenCode plugin compatibility.

Use when asked to create, scaffold, or write a new vibe-lab Agent Skill. Generates an Agent Skills compatible SKILL.md with optional scripts, references, assets, adapter symlinks, and verification.

Use when asked to create or design a vibe-lab workflow artifact that coordinates skills, agents, hooks, plugins, and verification gates across Claude Code, OpenCode, and OMP/Pi.

Use during implementation to run the smallest meaningful verification command after each focused change.

Use when a session is ending, context needs to be preserved for a future session, or when transferring work between agents. Generates a structured handoff document with open questions, next steps, and context summary.

Use when a bug report, error message, alert, or issue needs classification, deduplication, severity, ownership, refinement, and reusable triage learning before implementation.

Use when creating or modifying vibe-lab skills or adjacent artifact templates so Agent Skills compatibility, canonical source layout, adapter generation, and verification stay consistent.

Use when extracting actionable tasks from meeting notes, Slack threads, PR descriptions, specs, or other unstructured text. Converts ephemeral notes into tracked issues with context, acceptance criteria, deduplication, and learning capture.

Use during research or planning before implementation to choose a TDD mode, define red tests, preserve existing tests, and produce the test-first artifact for a feature, bugfix, or refactor.

Use when changing behavior so the agent drives the edit through a failing test, preserves existing tests, follows the selected TDD mode, and verifies red-green-refactor evidence.