verify-before-claim
Skill0 starsChanged 4 months agoMITView on GitHub
Before declaring work done, run it — do not conclude from a syntax check or a partial view. Trigger before saying "done", "fixed", "working", "ready", or reporting a result. Also trigger before quoting a number a tool produced, or claiming a tool is faulty. Prevents declaring success on the strength of parsing rather than running, testing the module instead of the entry point, and indicting a tool on a single reading channel.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- Mark393295827/graph-engineering-architecturestheir own version, first seen Nov 8, 2026
- hushh-labs/hushh-researchtheir own version, first seen Nov 1, 2026
- liushikuan63/Qoder-Optimizationtheir own version, first seen Nov 1, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- hushh-labs/hushh-research★ 40same name
- Mark393295827/graph-engineering-architectures★ 4same name
- liushikuan63/Qoder-Optimization★ 2same name
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 repository4 other files it gives its agents.
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Run it — even once, even trivially — before saying it's done.
- Distinctive: line 20
2. Test the entry point, not only the module: If you generated a package, import the package.
- Distinctive: line 21
If you wrote a CLI, invoke the CLI.
- Distinctive: line 23
3. A claim that indicts a tool needs a second channel: If a tool looks broken, confirm the evidence through a channel that tool did not …
- Distinctive: line 31
If I'm about to blame a tool: did I confirm through a different path?
- Distinctive: line 32
If I'm about to quote a number: did I check which source produced it and what it points at?
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Verify before you claim
Three measured failures, one session:
- A CLI passed a syntax check and raised on its first real call.
Path(".").parentisPath("."). - A generated
__init__.pyimported a symbol that no longer existed. Every module test passed; importing the package raisedImportError. No test imported the package. - Five files looked truncated through one channel and were intact through another. A bug report was one step from being filed against the wrong tool.
The three rules
1. Parsing is not running. A syntax check proves the code is well-formed, not that it works. Run it — even once, even trivially — before saying it's done.
2. Test the entry point, not only the module. If you generated a package, import the package. If you wrote a CLI, invoke the CLI. Unit-passing is not entry-point-working.
3. A claim that indicts a tool needs a second channel. If a tool looks broken, confirm the evidence through a channel that tool did not produce. Reading the same lie twice is not evidence. Corroboration is not repetition.
Before you say "done"
- Did I run it, or only check that it parses?
- Did I test what a user actually calls, or only an internal piece?
- If I'm about to blame a tool: did I confirm through a different path?
- If I'm about to quote a number: did I check which source produced it and what it points at?
The tell
"It should work now." Should is the word that precedes the bug. Run it and replace should with does — or with the actual error, which is more useful than a guess.
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.