Superpowers
obra/superpowersA 15-skill development methodology that makes coding agents clarify the goal, write a plan, work through it with subagents and verify before calling it done.
The repo-aware version of the grilling interview: it also builds a shared glossary and records hard decisions as ADRs while you talk.
The recommended starting point of the repo and heavily installed (about 1.25 million), but it only delegates and writes files into your repo.
grill-with-docs is a two-line skill that calls two others: grilling and domain-modeling. The result is the interview Matt Pocock recommends starting every change with, plus a paper trail. Grilling asks you decision questions in rounds with a recommended answer for each. Domain-modeling is the active discipline that challenges fuzzy or overloaded terms, invents edge-case scenarios and writes down what crystallizes: a GLOSSARY.md of project language (with a GLOSSARY-MAP.md when a repo has several contexts) and decision records under docs/adr.
The README argues that a shared language makes the agent use fewer words, name things consistently and think with fewer tokens, and calls this the strongest technique in the repo. Other skills then read the glossary and ADRs, for example tdd, wait-what and triage. The skill is marked disable-model-invocation, so you type /grill-with-docs yourself. Because the file only delegates, the sibling skills must be installed, and the result writes Markdown files into your repository, which you will want to review and commit like any other docs.
The SKILL.md is two lines calling grilling and domain-modeling; the substance sits in those two skills.
We scanned all 2 files in this skill folder and read every script. Nothing in it reaches beyond what the skill describes.
A one-line skill (user-invoked only) that tells the agent to call the grilling and domain-modeling skills. The only side effect comes from domain-modeling, which creates or edits GLOSSARY.md and ADR files in your repo; this skill itself has no scripts or network calls.
The skill folder (2 files) on Oct 29, 2026. We had every file and read every script.
NVIDIA SkillSpector static scan: 0 pattern hits. No SkillSpector hits.
A review, not a certification. Skills change after we check them — read the files before installing.
MIT — commercial use allowed (stated for the parent repository).
Not really. It is an interactive interview that needs your answers.
A 15-skill development methodology that makes coding agents clarify the goal, write a plan, work through it with subagents and verify before calling it done.
25 engineering-lifecycle skills and 9 slash commands (spec, plan, build, test, review, ship) that make coding agents follow senior-engineer workflows.
Every's plugin of 36 skills for a brainstorm, plan, build, review and capture-the-learning loop, with an optional hands-off /lfg pipeline.
A multi-harness plugin marketplace with 94 plugins, 202 agents, 185 skills and 105 commands for Claude Code, Codex, Cursor, OpenCode, Copilot, Antigravity and Pi.
It is Matt Pocock's interview skill for repos. It runs the grilling procedure together with domain-modeling, so the questions you answer also produce a GLOSSARY.md and architecture decision records.
Install the full mattpocock/skills set with claude plugin install mattpocock-skills@claude-plugins-official or npx skills@latest add mattpocock/skills, then run /setup-matt-pocock-skills once per repo and /grill-with-docs when starting a change.
It has no scripts or network calls. It does write Markdown documentation into your repository, so review those files before committing.
They run the same interview. grill-with-docs keeps a glossary and ADRs in the repo, while grill-me is stateless and suits work outside a repo.