opinion
Skill1 starsChanged 11 days agoMITView on GitHub
Asks workers on the current harness the same question and returns labeled takes. Default roster is the current model on every slot. Nobody writes project files. Optional --agents N (default 3, min 2, max 5). Apply when the user invokes opinion or /opinion, wants other takes, or says auto-pick. Use execute when this chat should do the work after the takes. Use deep-deliberation for checkpointed decisions. Use blinded-eval to A/B a skill change.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- noahcalibuso/herdr-fusiontheir own version, first seen Jul 23, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- noahcalibuso/herdr-fusion★ 1same 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 repository107 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/base.mdcCursor rule
- .cursor/rules/changelog.mdcCursor rule
- .cursor/rules/comments.mdcCursor rule
- .cursor/rules/release.mdcCursor rule
- .cursor/rules/seo-ai-search.mdcCursor rule
- .cursor/rules/testing.mdcCursor rule
- agent-readySkill
- authoring-skills-and-rulesSkill
- blinded-evalSkill
- cloakbrowser-fallbackSkill
- deep-deliberationSkill
- design-rationaleSkill
- dissectSkill
- documentation-writerSkill
- everything-seoSkill
- executeSkill
- generating-app-verifySkill
- inceptionSkill
- kit-workflowsSkill
- marketing-psychologySkill
- media-genSkill
- nlm-skillSkill
- proving-change-safetySkill
- refreshing-app-verifySkill
- release-deploySkill
- seo-auditSkill
- startup-marketing-brainSkill
- subsystem-walkthroughSkill
- trlSkill
- unmatched-workflowSkill
- agent-readySkill
- authoring-skills-and-rulesSkill
- blinded-evalSkill
- cloakbrowser-fallbackSkill
- deep-deliberationSkill
- design-rationaleSkill
- dissectSkill
- documentation-writerSkill
- everything-seoSkill
- executeSkill
- generating-app-verifySkill
- inceptionSkill
- kit-workflowsSkill
- marketing-psychologySkill
- media-genSkill
- nlm-skillSkill
- opinionSkill
- proving-change-safetySkill
- refreshing-app-verifySkill
- release-deploySkill
- seo-auditSkill
- startup-marketing-brainSkill
- subsystem-walkthroughSkill
- trlSkill
- unmatched-workflowSkill
- agent-readySkill
- authoring-skills-and-rulesSkill
- blinded-evalSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 6 are distinctive, and 5 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.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 46
Do not auto-start this skill.
- Distinctive: line 49
Do not invent a model name.
- Very specific: line 50
Never write a spawn identifier in a question, option, chat roster, worker prompt, or take label.
- Very specific: line 52
Do not write project files.
- Very specific: line 53
If this host cannot spawn a child at all, stop.
- Distinctive: line 77
If the user already named 2–5 models that exist on this host, use them.
- Very specific: line 81
If the user said auto-pick (or chose it), run auto-pick for agents families from references/hosts.md, announce with display names, then …
- Distinctive: line 101
Do not spawn yet.
- Very specific: line 103
If they assign each worker, ask one question per slot (not a multi-select).
- Distinctive: line 109
No structured question tool → numbered text fallback.
- Distinctive: line 123
Never invent a missing take.
≈ “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
Opinion
Leading words: current first, same-family ok, slug shield, auto-pick stays distinct, labeled takes, no writes.
Workers answer the same request. This chat does not edit the project.
User input
$ARGUMENTSOptional --agents N (also --n N, or a lone integer). Default 3.
Clamp to 2–5. Strip the flag from the question text. If the user already
named models (or current), use that list and ignore --agents. One
named model fills every slot.
Operating contract
- Do not auto-start this skill. The user must ask for it.
- Current-for-all is the default. Same family may occupy more than one slot when they chose that.
- Auto-pick is the only path that still requires distinct families.
- Do not invent a model name. Use the live list this harness exposes.
- Never write a spawn identifier in a question, option, chat roster, worker prompt, or take label. Map family → live-list name only in the spawn tool's
modelargument. Current →inherit(or omitmodel). - Workers are read-only. A worker that edits files has failed; discard that take.
- Do not write project files. If the user wanted the work done, use
execute. - If this host cannot spawn a child at all, stop. Auto-pick that cannot gather two families: offer Current-for-all or
deep-deliberation.
Runtime state
OPINION_STATE
host: cursor | claude-code | opencode | copilot | codex | other
agents: 2-5
roster: unset | Current,OpenAI,…
takes: 0/N
next_action: one action onlyPipeline
Copy these steps into the todo list verbatim. Skip with skip: <reason>.
- Parse
--agents(default 3, clamp 2–5). Detect the host. Read references/hosts.md. Build the live spawnable model list. Current-for-all does not need two families. - Resolve roster: models the user named, or Current for all, or auto-pick
agentsfamilies, or assign each worker. Announce with display names before spawning. - Spawn one read-only worker per roster slot. Same prompt. Parallel. Read references/worker-prompt.md.
- Render labeled takes (
[Current],[OpenAI], …). Stop. Do not merge into one verdict. Do not write files.
Roster
If the user already named 2–5 models that exist on this host, use them. Same family is allowed. If they named one model, or said current / inherit, fill every slot with that choice.
If the user said auto-pick (or chose it), run auto-pick for agents
families from references/hosts.md, announce with
display names, then spawn.
Otherwise ask:
prompt: Which models should answer?
options:
- Current model for all workers (Recommended)
- Auto-pick distinct families
- I'll assign each worker
- Say this in plain English
allow_multiple: falseOption ids: current, auto, assign, simplify. Never live-list names.
Say this in plain English is a meta question: explain Current vs
auto-pick with display-name examples (never spawn ids), then re-ask.
Do not spawn yet.
If they assign each worker, ask one question per slot (not a
multi-select). Each slot's first option is Current model (Recommended),
then one option per remaining family using display names and
fam-* ids from references/hosts.md. Same
family may be chosen on more than one slot.
No structured question tool → numbered text fallback.
Reply contract
Playbook: opinion
roster: Current, OpenAI, GrokThen the labeled takes. No files changed.
Delegate failure
- Retry one transient spawn failure.
- Never invent a missing take.
- Fewer than two successful takes → stop, ask whether to retry or abort.
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.

