claude-code-configs

Adopt Fable-class working discipline for this session — outcome-first communication, finish-the-turn autonomy, parallel tool use, faithful reporting. Use when the user says "/fable", "fable mode", "act like fable", or wants a smaller model (Opus, Sonnet) to work with the discipline of a Mythos-class model. Also invocable from other skills as a style preamble for long or high-stakes tasks.

Full workday recap across all tools — Jira, GitHub, Slack, Confluence, Google Docs, and Claude Code activity. Stores all data in SQLite with Turso sync for remote access. For Claude-only activity tracking, use /daily-review-v2 instead.

Claude Code activity tracking with SQLite/Turso storage — standup, weekly rollups, project reports, brag doc, friction tracking, goal tracking, heatmap, export. Data syncs to Turso for remote access. For full workday recap across all tools, use /daily-recap-v2 instead.

Transform a work-specific or company-specific skill/config into a generic, portable version. Replaces proprietary tools, internal URLs, and company references with generic equivalents while preserving the useful patterns and structure. Use when porting a work skill to the public config repo.

Sync ~/.claude/ global configs into this repo, sanitizing work-specific content (LinkedIn plugins, Captain MCP, work email, internal URLs). Use when the user wants to refresh this repo from their global Claude configs, pull in newly added skills/commands, or update sanitization rules. Two-pass flow with diff review before commit.

Deep discussion mode. No code changes, no file edits, no commands — pure thinking, exploration, and idea bouncing. Use when you want to explore a concept, design something, debate tradeoffs, or just think out loud. Claude will push back, ask questions, and present alternatives. Changes only happen if you explicitly say so.

Generate a smart commit message from staged or unstaged changes and create a commit. Use when the user wants to commit their work with a meaningful message.

Provide a deep, educational explanation of code. Use when the user wants to understand how code works, why it's designed a certain way, or learn about patterns and concepts used.

Proactive gap analysis on existing code. Finds what's missing — tests, error handling, observability, security, documentation, architectural holes — at micro (file), meso (module/feature), or macro (codebase) scope. Unlike /validate which checks recent changes, /gaps audits what already exists. Use when you want to know what's missing, undertested, unobserved, or architecturally incomplete.

Create a pull request with proper formatting. Use when the user asks to create a PR, make a pull request, open a PR, or wants to submit their changes for review.

Perform a thorough code review on a PR. Detects the language/framework, posts inline comments with severity labels (CRITICAL/MAJOR/MINOR/NIT/PRAISE) on specific lines, and leaves a single summary comment with overall verdict. Use when the user wants to review a PR, give feedback on a pull request, or do a code review.

Update an existing PR's title and description to reflect the latest commits on the branch. Use when the user has pushed new commits to an existing PR and wants the PR metadata to stay accurate.

Thoroughly validate recent changes through 6 independent analysis passes — implementation correctness, integration/side effects, adversarial failure analysis, observability, test quality, operational readiness — followed by a synthesis pass that cross-references all findings. Use after making changes to confirm they'll work as expected.

Deep documentation discovery across all available sources — Confluence/wiki, GitHub, JIRA, Slack, and semantic code search via Captain MCP. Fans out across multiple sources in parallel, ranks findings by authoritativeness, surfaces contradictions, and gives a verified/unverified/unknown verdict. Use when you need to validate an assumption, confirm a design decision, understand how something works internally, or back up a suggestion with real documentation. Can be invoked from other skills as a validation step.

Address PR comments intelligently by analyzing feedback, proposing fixes, and providing a summary of actions taken. Use when the user wants to respond to PR review comments.

Analyze all changes and split them into small, focused, reviewable commits grouped by logical concern. Use when the user wants to commit with best practices, create atomic commits, split changes into multiple commits, or make their PR easier to review.

Pull the latest changes from master/main into the current branch and resolve any merge conflicts. Use when the user wants to sync their branch with upstream, update from main, rebase on master, or fix merge conflicts.

Update all documentation inside the repository based on recent changes. Use when the user wants to refresh docs, update README, sync documentation with code changes, or maintain documentation.