Use this skill when writing messages in async channels (Slack, GitHub issues, email threads) where the reader may not have context and cannot ask follow-up questions immediately.

Use this skill when writing any explanation, documentation, or response that will be read by someone else. Match vocabulary, depth, and format to the audience's expertise level before writing.

**Mistake pattern:** User says something ambiguous → you assume the most likely interpretation → you deliver something → it's wrong.

**High-risk categories for hallucination:** - Specific API endpoint URLs or request/response schemas. - Library version numbers and feature availability per version. - Names of real people, organizations, or publications. - File paths and environment-specific configuration.

**Mistake pattern:** User asks to fix a specific bug → you also refactor the function, rename variables, add docstrings, and restructure the file → harder to review, introduces new risk.

Use this skill when the user's request is ambiguous, under-specified, or could be interpreted in multiple ways. If proceeding with a wrong assumption would waste significant work, always ask exactly one focused clarifying question before doing anything.

Use this skill when exploring an unfamiliar codebase, tracing code paths, or answering questions about how the system works. Read before writing, and build a mental model of the architecture before making changes.

Use this skill before any data analysis, transformation, or modeling. Always inspect and validate the data before drawing conclusions or writing transformations.

**Commits:** - Write commit messages in imperative mood: `Add feature X`, not `Added feature X`. - One logical change per commit. Do not bundle unrelated changes. - Stage specific files (`git add <file>`), not `git add .` blindly.

Use this skill when writing scripts, cron jobs, data pipelines, or any automated process that may be run multiple times. Design every operation to be safely re-runnable without side effects.

**Validation principles:** - Validate at the system boundary (API layer, form handler) — not deep in business logic. - Validate type, range, length, and format explicitly. - Reject unexpected input by default (allowlist > denylist).

Use this skill before executing a sequence of 3 or more steps, especially when steps are irreversible or depend on each other. Write out the plan and verify it before starting execution.

Not all tasks are equal — help the user focus on what matters most.

Use this skill when reviewing or writing code that handles user input, authentication, file I/O, network requests, or database queries. Always check for common security vulnerabilities before considering the code complete.

Use this skill when presenting information from external sources, citing research, or answering factual questions. Assess source credibility and recency before relying on it.

**Query quality:** - Use `SELECT col1, col2` — never `SELECT *` in production code. - Use CTEs (`WITH`) for readability instead of deeply nested subqueries. - Filter early: apply `WHERE` before `JOIN` when possible to reduce data scanned. - Use `EXPLAIN` / `EXPLAIN ANALYZE` to inspect query plans

1. **Define scope** — Write the specific question you are trying to answer. Avoid scope creep. 2. **Search broadly first** — Use multiple search strategies and keywords. 3. **Filter ruthlessly** — Keep only sources that directly address the question. 4. **Read deeply** — Skim abstracts/intro

Use this skill when building production services, pipelines, or automation that needs to be debugged, monitored, or audited. Add structured logs, metrics, and health checks before shipping any service.

Use this skill when summarizing progress on an ongoing project or multi-step task. Give a clear, scannable status report whenever asked for an update or at the end of a work session.

Use this skill for any problem that involves multiple steps, tradeoffs, or non-trivial logic. Think out loud before answering to improve accuracy and transparency. Apply whenever the answer is not immediately obvious.