Personal AI agent guides, templates, and learnings
This repository also configures its own agents. See what safeword tells them →
Cursor rule Cursor
USE WHEN starting feature work, running /bdd, resuming a BDD ticket, or user says 'add', 'implement', 'build', 'feature', 'resume', 'continue'. Orchestrates BDD phases.
Cursor rule Cursor
USE WHEN in BDD intake phase OR ticket has phase:intake. Guides discovery and context gathering for features.
Cursor rule Cursor
USE WHEN in BDD done phase (verify.md exists). Close the ticket — verification already happened in verify.
Cursor rule Cursor
USE WHEN in BDD plan-implementation phase. Author impl-plan.md, run figure-it-out on load-bearing choices, ADR lifecycle, review before implement.
Cursor rule Cursor
USE WHEN in BDD define-behavior or scenario-gate phase. Guides Given/When/Then scenario creation and validation.
Cursor rule Cursor
USE WHEN BDD thresholds exceeded (2+ stories, >15 scenarios, >20 tasks). Split protocol and examples.
Cursor rule Cursor
USE WHEN in BDD implement phase. RED/GREEN TDD cycle; run /refactor after GREEN.
Cursor rule Cursor
USE WHEN in BDD verify phase OR all scenarios marked [x]. Evidence gate — cross-scenario refactor, then /verify and /audit.
Cursor rule Cursor
Collaborative brainstorming and rubber-ducking — divergence-first thinking partner. Use when the user wants to explore options, weigh approaches, or think through uncertainty before committing to a direction ('brainstorm', 'rubber duck', 'help me think', 'explore options', 'what are the tradeoffs').
Cursor rule Cursor
@.safeword/SAFEWORD.md.
Cursor rule Cursor
Finds the real cause of a bug, in four steps, before you touch a fix. Use when encountering bugs, test failures, unexpected behavior, or when previous fix attempts failed. Enforces investigate-first discipline ('debug this', 'fix this error', 'test is failing', 'not working').
Cursor rule Cursor
Extract tacit knowledge through non-obvious microquestions — things only the user knows that can't be found in code, docs, or research. Use when about to guess at intent, context, or constraints, or when user says 'ask me', 'what do you need to know', 'what's missing'. Skips questions answerable by reading the…
Cursor rule Cursor
Explore and debate options with fresh documentation and research before committing. Use when facing a real decision with multiple plausible approaches — library/framework choice, architecture call, API or schema design, algorithm selection. Looks up current docs and recent research, weighs options on correctness and…
Cursor rule Cursor
Internal fallback used only right after the shared review coordinator reports it's run out of reviewer routes. Not something a user invokes directly.
Cursor rule Cursor
Deep quality review of any work-product — code, docs, specs, plans, decisions — with web research. Use when explicitly verifying against latest docs ('double check against latest', 'verify versions', 'check security'), pressure-testing correctness and elegance, or needing deeper analysis beyond the automatic hook.…
Cursor rule Cursor
Tests whether a proposed product bet has credible demand. Use when Why now depends on an unresolved, decision-critical demand claim or when the user explicitly asks for demand research. Do NOT use for child features, mandated work, parity work, or when a cheaper experiment can answer the question first.
Cursor rule Cursor
Prepare a pull request for human review and decide whether it may leave Draft. Use when creating or rewriting a PR description, marking a PR ready, responding to review, or checking mergeability.
Cursor rule Cursor
Systematic refactoring with small-step discipline. Use when user says 'refactor', 'clean up', 'restructure', 'extract', 'rename', 'simplify', or mentions code smells. Enforces one change → test → commit when the commit can stay scoped. For structural improvements, NOT style/formatting (use /lint). NOT for adding…
Cursor rule Cursor
Files Safeword's sanitized spooled retrospective drafts to its upstream tracker. Use only when a trusted Safeword Stop continuation or authenticated closeout cleanup guard output names a spool path. Do not use for ordinary retros, project issues, or user-authored drafts.
Cursor rule Cursor
Step-aware quality review at TDD phase boundaries — review test quality after RED, implementation correctness after GREEN, completed scenario after REFACTOR. Use when user says 'review my test', 'review my implementation', 'is this GREEN solid?', or just finished a TDD step. (Note: Claude Code triggers this…
Cursor rule Cursor
How to write good tests. Use when writing tests in any context — 'write tests', 'add tests', 'test this', 'need tests for', 'improve coverage'. Also consult when writing tests during BDD or debugging. Core knowledge for test quality across all workflows.
Cursor rule Cursor
Ticket system and work logs for context anchoring during complex work. Use when creating tickets, managing work logs, referencing ticket IDs, or when work needs context anchoring (multi-step tasks, debugging, investigation).
Cursor rule Cursor
USE WHEN in BDD plan-execution phase. Turn the accepted Implementation Plan into startable work and review before coding.
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: