DeepReason: Skill for Claude Code

.claude/skills/dr-verify-outcome/SKILL.md

dr-verify-outcome is a skill for Claude Code from AHepi/DeepReason. It costs 44 tokens per session (881 once invoked), scanned A, original, MIT.

A verification step that checks whether a DeepReason fix met the success condition written in GOAL.md, then records a PASS or FAIL in VERIFY.md.

In plain words
What is it for?
Run the required success command, recheck affected readers or validators against known failing and working inputs, and optionally perform one guarded live run when the goal requires it.
Why use it?
It prevents fixes from being called successful without evidence. It also keeps verification separate from changing code.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions CLAUDE.md.

This is AHepi/DeepReason's own configuration. It tells Claude Code how to work on DeepReason itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything DeepReason configures →

Reuse

Borrowing it

Nothing to install: this file belongs to AHepi/DeepReason. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/AHepi/DeepReason/main/.claude/skills/dr-verify-outcome/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/AHepi/DeepReason

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for dr-verify-outcome

README.md
[![agentmods](https://agentmods.dev/badge/skills/ahepi/deepreason/dr-verify-outcome.svg)](https://agentmods.dev/skills/ahepi/deepreason/dr-verify-outcome)
Your own site
<a href="https://agentmods.dev/skills/ahepi/deepreason/dr-verify-outcome"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-verify-outcome.svg" alt="Measured on agentmods" height="20"></a>
Per session 44 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 881 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00044 $0.00881
Opus 5 $0.00022 $0.00441
Sonnet 5 $0.00009 $0.00176
Haiku 4.5 $0.00004 $0.00088

Measured 6d ago against content hash 96c320c4b54c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

dr-verify-outcome scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/dr-verify-outcome/SKILL.md · 74 lines

How it starts

The opening of the file, as written. The whole thing — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Verify the outcome

Input: GOAL.md's machine-decidable success criterion + the pushed fix. Output: VERIFY.md with verdict PASS or FAIL. You verify; you do not patch. Any new code change belongs to a fresh cycle through the orchestrator.

Ladder of proof (ascend only as far as the goal requires)

  1. Criterion command. Run GOAL.md's success command verbatim. Paste output.
  2. Historical roots. If the fix changed a reader/validator, re-run verify_root against the committed roots that showed the defect AND one known-good root: the target violations must disappear, everything else must be unchanged. List remaining violations by class — do not summarize them away.
  3. Live run (only if GOAL.md demands live proof). At most ONE attempt, fully guarded:
    • Preflight: env file present, deepreason importable, run identity free (retire + commit rename first if occupied).
    • Launch detached from the ladder's own directory: setsid nohup ./<ladder>.sh & disown.
    • Arm rollback insurance: the snapshot loop (commit+push the experiment dir every ~5 min) and a monitor on progress.jsonl state/phase + the driver log's rc= lines, alerting on both success AND failure signatures.
    • Judge only typed outcomes: run state, stop_reason, audit JSON, verify_root, FINDINGS.md. Model prose is not verification.
  4. Know what a live run can and cannot prove. Model behavior is stochastic across identically-configured runs (a capability channel used in one attempt may go unused in the next). A live attempt that never reaches the fixed path is INCONCLUSIVE for that path — say so; the offline regression remains the proof of correctness. Do not burn repeated live attempts chasing a stochastic path: one relaunch maximum, then record the residue.

VERIFY.md template

# Verification
Criterion command + output: <pasted>
Historical roots re-checked: <root -> before/after violation classes>
Live attempt (if any): <run id, typed stop, tokens, one-line audit>
Verdict: PASS | FAIL | PASS-offline/INCONCLUSIVE-live
Residue (honest): <what remains unproven or parked, or "none">
Errata: <docs/ERRATA.md entry id(s) added, or "errata: none">

Read the full file on GitHub · 74 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 74 lines · 44 tokens per session scan A 96c320c4b54c

Subscribe to this mod's changes

dr-verify-outcome is a skill published in the GitHub repository AHepi/DeepReason (142 stars, last pushed today), licensed MIT. It adds 44 tokens to every session and 881 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens