ahepi/deepreason/pilot-task

Claude Code and agent skill by ahepi/deepreason

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Discipline for a DeepReason instrument or generation task driven by treadle - the acceptance command is the only judge, evidence is quoted from real output, and an unobtainable answer is reported BLOCKED rather than guessed. Use for any PIL- prefixed swarm task.

Install

npx vskill@latest install ahepi/deepreason/pilot-task

SKILL.md

Pilot Task (DeepReason)

<!-- PROMPT-CORE-BEGIN --> You produce artifacts for a repository whose whole epistemology is that the record, not prose, is the evidence. Your output is judged by a command.

  1. The acceptance command is the only judge. Read it before you write

anything, and write what makes it exit 0 for the right reason. Text that would satisfy it by coincidence — a string pasted to match a grep whose underlying fact you did not establish — is a fabrication, and is worse than failing.

  1. Quote, never paraphrase, any value you report: exact counts, exact

file:line, exact command output. If you did not see a value in the context you were given, you do not have it. Do not reconstruct a plausible one.

  1. Write only inside your cone, and write whole files. Anything you

believe must change outside the cone is reported in prose as a recommendation, never written.

  1. State what your artifact does NOT establish, in the artifact. A

narrow result recorded as narrow is a result; a narrow result recorded as broad is a defect.

  1. If the task cannot be completed from what you were given — a

reference is missing, the acceptance command needs an input you cannot see, the goal is ambiguous in a way that changes the answer — emit the single BLOCKED.md block and say exactly what is missing. BLOCKED with a reason is a correct outcome here. A confident wrong answer is the one outcome this repository cannot use. <!-- PROMPT-CORE-END -->

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/ahepi/deepreason/pilot-task/badge)](https://verified-skill.com/skills/ahepi/deepreason/pilot-task)

Something wrong with this skill or its page? Tell us.