cold-read
Skill14 starsChanged 5 days agoMITView on GitHub
Use when dispatched to cold-read a plan, as an agent that was not in the conversation that wrote it, to say whether it can be carried out from the document alone. Produces what the reader would build, its questions, and a verdict, ready or not ready, repeated as the writer fixes the plan. Not for judging the design (interview, pr-review), not for the mechanical checks (aiview plan check, tracker check), and not for writing the plan (write-plan).
6 other projects use a version of this.
Adopted by
Most recently taken up by
- Thakay/agent-skillstheir own version, first seen Nov 1, 2026
- FiredMosquito831/my-writing-companiontheir own version, first seen Nov 1, 2026
- me2resh/cold-readtheir own version, first seen Oct 25, 2026
- Ovich/swarmstormword for word, first seen Oct 25, 2026
- scottlivingston/claude-skillstheir own version, first seen Oct 25, 2026
- kevmoo/kevmoo_skillstheir own version, first seen Oct 13, 2026
Other versions6 projects use a version of this.
Word for word · 1
- Ovich/swarmstorm★ 14100% same
Edited · 0
None found.
Wrote their own version · 5
- kevmoo/kevmoo_skills★ 22same name
- FiredMosquito831/my-writing-companion★ 2same name
- scottlivingston/claude-skills★ 1same name
- me2resh/cold-read★ 0same name
- Thakay/agent-skills★ 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 repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- brainstormSkill
- brainswarmSkill
- interviewSkill
- whiteboard-defenseSkill
- complexity-reworkSkill
- deepen-moduleSkill
- dependency-graphSkill
- simplify-reworkSkill
- simplifySkill
- write-diagramsSkill
- aiviewSkill
- swarm-boardSkill
- execute-planSkill
- long-horizonSkill
- orchestrate-planSkill
- guide-complexitySkill
- guide-createSkill
- guide-deepen-moduleSkill
- helpSkill
- next-stepSkill
- orchestrateSkill
- write-planSkill
- roadmapSkill
- technical-writingSkill
- write-conventionsSkill
- pr-reviewSkill
- recurring-review-findingsSkill
- swarmSkill
- design-prototypeSkill
- point-atSkill
- validate-deliverySkill
- verification-skill-createSkill
- create-workbenchSkill
The file
We split this skill into the rules it gives an agent
1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
1
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 23
Talk it through with the writer: Ask back, push on an answer that does not convince you, say how you read a passage and let them correct …
“Common” is measured: kept by at least 6 in 10 of the 4 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
Cold read
You stand for the agent who will carry this plan out with nothing but the document. You were not in the conversation that wrote it, and that is the point: its writer cannot see what is in their head and not on the page. Your job is to find it. A question you guess past is a task built on a guess.
What to do
- Read the plan through the
aiviewskill, every document it links, and the code it names where you need it to understand a part. Do not edit anything. - Answer, in this order:
- What you would build: the goal, the design and the phases, in your own words, short. The writer compares it with what they meant.
- Your questions: everything you would have to ask, or guess, before writing any phase's tasks. Each one says where in the plan it arises.
- Your verdict:
readywhen no question is left,not readyotherwise.
- Talk it through with the writer. Ask back, push on an answer that does not convince you, say how you read a passage and let them correct you, propose a wording. The exchange is how you both find what the plan leaves unsaid; keep it going until you understand the work as they do.
- A question closes in the plan, not in the talk. Once you agree, re-read the changed place: what the next agent needs and only the conversation holds is still a question. Give the verdict again with what is closed and any new question the change raised.
What to check
- Every term is defined or plainly the codebase's.
- Every decision has its reason, so a case it did not foresee can be decided the same way.
- The plan agrees with itself: the design, the decisions, the phases and the tracker say the same thing.
- Every done-when is checkable: a command, an observable result, never a judgement.
- Every file, module and document named exists, or the plan says it is created.
- Each phase's tasks can be written from the plan alone: what it builds on, what it must not touch.
You judge clarity, not the design. A design you would have done otherwise is not a question; a design you cannot understand is. A flaw you cannot leave unsaid is named once, as a flaw, apart from the questions.
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.