blast-radius
Skill42 starsChanged 40 days agoApache-2.0View on GitHub
Before review sign-off and before merge, find what a diff could break OUTSIDE the edited files, and prove the one fact that makes it safe by running real code or a query. Use for "what could this break", for any PR touching shared code, SQL, migrations, env or jobs, and for small diffs you don't trust. The result goes in the PR.
- Reads credentials
111 other projects use a version of this, 6 of them edited.
Adopted by
Most recently taken up by
- marmot-protocol/mdktheir own version, first seen Nov 9, 2026
- Yeachan-Heo/oh-my-claude-sisyphustheir own version, first seen Nov 8, 2026
- Loner1024/skillstheir own version, first seen Nov 8, 2026
- imranbarbhuiya/skillstheir own version, first seen Nov 8, 2026
- caidanw/skillstheir own version, first seen Nov 8, 2026
- pedrohcgs/claude-code-my-workflowtheir own version, first seen Nov 8, 2026
Used with
Projects with it also use ketchup-video, principle-never-block-on-the-human and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
tristanphvnSkillketchup-video
Catch the user up like /keel:ketchup, with a short silent video of slides for what happened and the waiting calls as text under it. Use for /keel:ketchup-video.
Also in 14 projects that use this
jaydubya818Skillprinciple-never-block-on-the-human
Apply when safe reversible implementation details can proceed inside scope.
Also in 44 projects that use this
ianclearySkillprinciple-explain-the-number
Apply before you trust, report, or act on a number you measured: a speedup, a regression, a throughput, a latency, or an eval result. Find what limits it, and rule out that it measured something other than the work you think.
Also in 29 projects that use this
eryzerzSkillmake-bot-ui
Use when building a custom UI (page, dashboard, buttons) that should wake a Grok Bot over a webhook, when the user must provide a webhook sender key, or when exposing that UI on Tailscale.
Also in 20 projects that use this
bnemaSkillsetup-pstack
Configure which Pi models pstack uses per role. Detects available provider/model specs and writes pstack's user configuration. Use for /setup-pstack, "configure pstack models", or changing pstack model choices.
Also in 25 projects that use this
jaydubya818Skillprinciple-make-operations-idempotent
Apply when an operation must tolerate retries after partial completion.
Also in 46 projects that use this
- +6 more from jaydubya818
Other versions111 projects use a version of this.
Word for word · 0
None found.
Edited · 6
- savannahfeder/agentbox★ 37612% same
- krishhgg/t3term★ 1111% same
- educlopez/mallard★ 413% same
- stainless-code/lossless★ 410% same
- arusli1/askills★ 110% same
- antoniowolkk/junior-agent★ 015% same
Wrote their own version · 105
- Yeachan-Heo/oh-my-claude-sisyphus★ 41ksame name
- Yeachan-Heo/oh-my-claudecode★ 41ksame name
- backnotprop/pstack★ 3.8ksame name
- pedrohcgs/claude-code-my-workflow★ 1.7ksame name
- pixeltable/pixeltable★ 1.7ksame name
- moeru-ai/xsai★ 667same name
- jtaroreh/agystack★ 128same name
- uzairansaruzi/p3-stack★ 126same name
- Dimon94/dverity★ 107same name
- Dimon94/skills★ 107same name
95 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- starter-kit/AGENTS.mdAGENTS.md
- agent-stackSkill
- bug-review-boardSkill
- openrig-project-setupSkill
- project-onboardingSkill
- review-lensesSkill
- technical-writingSkill
- unslopSkill
- verification-guideSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.
0
Common
Most of the 106 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 28
Never invent a caller or an API.
- Very specific: line 29
Prove the one fact: Write a small script, test or read-only query that exercises the real code or data, run it, and paste the output.
- Very specific: line 54
Write it with the unslop skill.
“Common” is measured: kept by at least 6 in 10 of the 106 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Blast radius
Listing callers is not the job; rg does that in a second. The job is the breakage a search won't show you.
When
- The reviewer runs it before signing off a PR that touches shared code, SQL or queries, migrations, env or config, background jobs, a public API or a data format.
- The integrator checks that the PR carries it before the merge gate. For risky-tier PRs it is required.
Steps
- Read the change. The diff, the symbols it adds, changes and deletes, and what the code now does differently, including what the diff doesn't spell out.
- Find the one fact it is safe because of. Most risky-looking changes are safe because of one fact, like "this migration only adds a nullable column" or "this helper is only called from the admin page". If that fact holds, most risks go away at once. Spend your time here.
- Look where search stops. A database column read by another service or a report. JSON that a client or a stored job parses. An env var another deploy target sets differently. A cron job or queue consumer that runs later. A library's pinned version or local patch. Code three calls downstream.
- Rate each risk. How it breaks, the real
file:line, how likely, how bad, and how to check. Keep confirmed risks separate from the ones you checked and cleared. A search that finds nothing is an answer; say what you searched. Never invent a caller or an API. - Prove the one fact. Write a small script, test or read-only query that exercises the real code or data, run it,
and paste the output. Queries go against the development database only (
.env.local), never production.
How sure you are
Say how far the safety fact got:
- You said so. This proves nothing.
- You pointed at the line (
file:line, or the library's own source). - You walked the failure case step by step and showed it can't happen.
- You ran code that calls the real function and fails loudly if you are wrong.
- You reproduced it in the running app.
Aim for 4. If you stop earlier, write "unproven" next to the fact.
What goes in the PR
A ## Blast radius comment or section:
- What changed, including the part that isn't obvious.
- Safe because: the one fact, the level it reached, and the proof output.
- Risks: each with how it breaks,
file:line, likelihood, cost and how to check. - Cleared: what you checked and why it is fine.
- Before merge: the cheapest test or repro that would catch the real bug.
Write it with the unslop skill. No secrets or production data in the comment.
Adapted from blast-radius in pstack by Lauren Tan (v0.15.5,
commit fae2c6e, 2026-09-29), MIT licence; the licence text is in LICENSE beside this file. Changes: tied to our
reviewer and integrator steps, the result posted on the PR, queries limited to the development database; references
to pstack's how, why and arena skills removed.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

