ground
Skill0 starsChanged 39 days agoMITView on GitHub
Establish the smallest evidence-backed model of current reality sufficient to proceed. Use when entering an unfamiliar system, current behavior or constraints are uncertain, assumptions matter, or before consequential changes.
7 other projects use a version of this.
Adopted by
Most recently taken up by
- synaptiai/agent-capability-standardtheir own version, first seen Nov 1, 2026
- halakou/groundtheir own version, first seen Oct 20, 2026
- ilanbm/kpoppertheir own version, first seen Sep 23, 2026
- envisioning/skillstheir own version, first seen Sep 22, 2026
- reutavidan/zetesis-plugintheir own version, first seen Aug 6, 2026
- ahelme/skillstheir own version, first seen Jul 5, 2026
Other versions7 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 7
- ilanbm/kpopper★ 12same name
- synaptiai/agent-capability-standard★ 6same name
- envisioning/skills★ 2same name
- ahelme/skills★ 1same name
- yiyaw-lab/agent-armor★ 1same name
- halakou/ground★ 0same name
- reutavidan/zetesis-plugin★ 0same 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 repository16 other files it gives its agents.
- create-checkSkill
- decideSkill
- design-changeSkill
- design-moduleSkill
- harden-boundarySkill
- harden-checkSkill
- intentSkill
- investigateSkill
- learnSkill
- pave-pathSkill
- ratchetSkill
- realizeSkill
- specifySkill
- validateSkill
- verifySkill
- wire-checkSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 7 edited versions we read keep these.
3
Distinctive
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 13
Inspect primary evidence: Prefer the running system, source, tests, configuration, history, logs, traces, persisted state, and …
- Distinctive: line 15
Classify claims: Keep observed facts, supported conclusions, assumptions, and unknowns distinct.
- Distinctive: line 28
Stop and correct course if you are treating documentation as runtime truth, silently converting inference to fact, exploring unrelated …
“Common” is measured: kept by at least 6 in 10 of the 7 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
Ground
Establish what is actually true before reasoning from it. Do not propose changes unless the task explicitly requires continuing beyond grounding.
Work
- Frame the question. State what must be understood and which facts could change the next decision.
- Inspect primary evidence. Prefer the running system, source, tests, configuration, history, logs, traces, persisted state, and authoritative records over summaries. Follow the relevant path end-to-end when behavior matters.
- Reconstruct current state. Explain the relevant actors, boundaries, control/data flow, state changes, external effects, and constraints. Run or probe the system when static inspection cannot establish behavior.
- Classify claims. Keep observed facts, supported conclusions, assumptions, and unknowns distinct. Surface contradictions rather than resolving them by intuition.
- Bound the search. Investigate only gaps that could materially change the next move. Stop at decision sufficiency.
Return
- Current state — the smallest useful model.
- Evidence — concrete sources/observations supporting material claims.
- Constraints — relevant boundaries that already exist.
- Unknowns — material unresolved questions.
- Implications — what this state means for the next move, without smuggling in an implementation decision.
Failure signals
Stop and correct course if you are treating documentation as runtime truth, silently converting inference to fact, exploring unrelated areas, or gathering information without reducing decision-relevant uncertainty.
Done when
Material questions are answered with evidence or explicitly marked unknown; contradictions are visible; and more grounding is unlikely to change the next consequential move.
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.