semantic-decision
Skill0 starsChanged 2 days agoMITView on GitHub
Make an explicit, structured judgment from evidence and bounded choices. Use for task sizing, skill routing, orchestration need, test impact, review findings, recovery, decision tiers, or whether human/deeper review is warranted. Defaults to the current LLM; configured Cloudflare Clef/Clef-flash can replace the provider.
In this repository30 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- llms.txtllms.txt
- agent-first-docsSkill
- analytics-reviewSkill
- architectSkill
- autopilotSkill
- browseSkill
- create-prSkill
- design-uiSkill
- foremanSkill
- harness-auditSkill
- humanizer-viSkill
- image-genSkill
- implementSkill
- investigateSkill
- marketing-videoSkill
- perf-auditSkill
- plan-draftSkill
- qaSkill
- reconSkill
- release-prdSkill
- releaseSkill
- review-prSkill
- setup-projectSkill
- teamSkill
- telemetry-reviewSkill
- testSkill
- token-auditSkill
- triageSkill
- ui-auditSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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 13
Read the contract. Take the caller's evidence, decision kind, bounded choices and applicable rubric from context or files.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
semantic-decision
A reusable step inside the work, independent of its provider. Follow the
Auto-Decision Framework; mechanical
facts need no model, and User Challenges retain their escalation channel.
Shared refs are filesystem paths beside this skill's directory, so read them
by path, not as skill://.
- Read the contract. Take the caller's
evidence, decision kind, bounded choices and applicable rubric from context
or files. Keep cited evidence and criteria in a request JSON file; no ticket
is required. Missing evidence: gather it or return
NEEDS_CONTEXTonly when the fact cannot be derived, never silently invent it. - Run
bbs semantic-decision --kind <kind> --input <request.json>.NEEDS_LLMis the normal default provider: make the judgment in this session.FALLBACK_LLMuses the same path after a Cloudflare failure or restriction. Record the LLM's choice and cited reason per question with--llm-answers <answers.json>against the same request. Never launch a second model for the LLM provider or alter user config to complete the step. - Reconcile
DECIDEDanswers with evidence and the consumer's mandatory floors. If a Cloudflare answer conflicts, record an LLM resolution with the discrepancy. Keep the structured result and evidence with the caller's artifact; return choices to the caller, which owns execution and permission.
Output: STATUS: DONE, VERDICT: DECIDED, PROVIDER: llm|cloudflare,
DECISIONS: <result JSON/path>, EVIDENCE: <request and cited rationale>.
An unresolved User Challenge uses the preamble's NEEDS_CONTEXT channel.
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.

