Agent skill

Dr Execute Step

by AHepi in AHepi/DeepReason

Execute exactly one unchecked step from CHECKLIST.md, prove its done-criterion, record the output, and stop.

MITAuto-check passed

Install Dr Execute Step

skills CLI
$ npx skills add AHepi/DeepReason --skill dr-execute-step -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install AHepi/DeepReason dr-execute-step --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/AHepi/DeepReason.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/dr-execute-step .claude/skills/dr-execute-step && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
dr-execute-step
GitHub stars
141
Token cost
~1.9k tokens
SKILL.md length
1,063 words
Files
1
Skills in repo
30
Repo updated
First seen
Licence
MIT

At a glance

Execute exactly one unchecked step from CHECKLIST.md, prove its done-criterion, record the output, and stop.

  • Works in 6 steps: Re-read REQUEST.md (including… → Confirm the step still makes sense… → Execute the action. Only files this… → …
  • SKILL.md covers Procedure, Map obligations (docs/map/), Durable tests, checks, and… and Style discipline for code steps, plus 1 more section
  • Calls git and python

What it does

Dr Execute Step is an agent skill from AHepi/DeepReason. Execute exactly one unchecked step from CHECKLIST.md, prove its done-criterion, record the output, and stop. The only skill in the change workflow allowed to modify the tree. Invoke repeatedly, once per step.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The licence is MIT.

Example prompts

  • “/dr-execute-step”

Requirements

  • Python 3

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

  1. Re-read REQUEST.md (including Amendments) and CHECKLIST.md in
  2. Confirm the step still makes sense against the tree (a prior step
  3. Execute the action. Only files this step's spec item names may
  4. Run the done-criterion command. Paste its real output (trimmed to
  5. **If this step changed behaviour, update the map in the SAME
  6. Mark the box, update CHECKLIST.md — including its header State

What it can do on your machine

Read from SKILL.md and the folder at commit 9607fba. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • python

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Dr Execute Step loads about 1.9k tokens when it runs. Until then it costs about 56 tokens; SKILL.md has 1,063 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~56
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from AHepi/DeepReason at commit 9607fba, republished under its MIT licence (© AHepi). 1,063 words, ~1,914 tokens.

Download SKILL.mdSave it as .claude/skills/dr-execute-step/SKILL.md (or your agent's skills folder).
name
dr-execute-step
description
Execute exactly one unchecked step from CHECKLIST.md, prove its done-criterion, record the output, and stop. The only skill in the change workflow allowed to modify the tree. Invoke repeatedly, once per step.

Execute one step

Input: CHECKLIST.md. Output: one more checked step with its done-criterion output pasted beneath it. You do this for ONE step, then return. The loop lives in the orchestrator, not in you — that is what keeps a long change from drifting.

Procedure

  1. Re-read REQUEST.md (including Amendments) and CHECKLIST.md in full. Find the FIRST unchecked step. That is your entire job. Do not read ahead "to be efficient"; do not batch steps.

  2. Confirm the step still makes sense against the tree (a prior step may have failed silently). If the tree contradicts the step — file missing, test already passing, root identity occupied — do not improvise: record the contradiction under the step, commit, and return to the orchestrator (route: dr-plan-steps).

  3. Execute the action. Only files this step's spec item names may change. Mid-step discoveries ("this file also needs...") go to PARKED.md or, if the change cannot land without them, back through dr-spec-change as an amendment — never just typed in.

  4. Run the done-criterion command. Paste its real output (trimmed to the relevant lines) under the step. If it does not match expected: the step is NOT done — leave it unchecked, record the output and one line on the mismatch, and return to the orchestrator. Two failures of the same step = stop condition, in the standard format — canonical in dr-drive-harness §6's calibration note.

  5. If this step changed behaviour, update the map in the SAME commit — see "Map obligations" below. If it changed the packaging surface (pyproject entry points, CLI commands, MCP tools/schema, wheel layout), update scripts/wheel_smoke.py's pinned expectations and re-run the smoke in the same commit too — no gate runs it for you.

  6. Mark the box, update CHECKLIST.md — including its header State: line (next step, blockers), which is what a fresh session resumes from — and if the step is tagged [COMMIT] (or changed any file): git add this step's files, then run python tools/diff_budget.py <tranche-base> --ceiling <SPEC.md's ceiling> --paths <SPEC.md's declared areas> and read its DIFF_BUDGET_RESULT_V1.verdict. WITHIN/NO_CEILING: continue. EXCEEDED is a STOP in the standard format (decision, priced options, recommendation), not a footnote — an estimate-only ceiling trips on nothing unless read. Alongside it, run python tools/blast_radius.py --files <this step's actually git-added files> --symbols <this step's actually touched top-level defs, from the diff hunks> --against <tranche-base> (Rung G6, docs/map/INV-frozen-surfaces.md) and diff its frozen_surface_contacts/reachability output against THIS document's own Frozen-surface contact forecast and Blast-radius census sections in SPEC.md. Any frozen_surface_contacts entry not already named in SPEC.md, or any reachability entry whose direction is newly_dead/newly_live and was not predicted, is DRIFT — a STOP in the exact same format as diff_budget.py's own EXCEEDED, never a footnote (docs/ERRATA_EXECUTOR.md, "the frozen-surface stop did not hold"). No drift: continue. Then commit and push now.

     git add <files this step touched> <map files> <tranche-dir>
     git commit -m "step <n>: <checklist line>"
     git push -u origin <branch>   # retry x4, backoff 2s 4s 8s 16s

Map obligations (docs/map/)

The map is part of the change, not a chore after it.

  • A step that changes what a caller may do, what a guard admits, or where a rule is enforced, updates the covering SUB-/CON-/SEAM- document in the same commit.
  • A step that changes an interaction updates the SEAM- document before the subsystem ones — the seam is what the next reader opens first, and a correct pair of subsystem docs with a stale seam between them is worse than either being stale alone. The file is docs/map/SEAM-<a>-x-<b>.md (sides alphabetical); how to change one is docs/map/REC-change-a-seam.md; how to write one is docs/map/SCHEMA.md.
  • New behaviour needs a new check at column 0 that would fail if the behaviour regressed. Run it before you write it down.
  • Advance Verified-at: only if you re-ran that document's checks.
  • python tools/docs_verify.py must pass before you commit; a failure is a failed step, exactly like a failed test.
  • A step that only writes tests or records evidence changes no map document. Do not touch stamps you did not verify.
Show full SKILL.md (409 more words)Show less

Durable tests, checks, and probes

Anything you add here must survive dramatic repo changes — refactors, renames, reformats — failing only when the CLAIM it guards stops being true. Five rules, each paid for once already:

  1. Pin to committed, immutable evidence. A test or check may open only roots and fixtures that git ls-files knows; regression tests name their motivating run in the docstring. Session-local artifacts die with the session and take the check's meaning with them (docs/ERRATA.md E7).
  2. Anchor to meaning, not form. Prefer behavior (call the function, compare typed outcomes), structure (AST shape, resolved-call counts), or counts over literal source text. When a textual marker is unavoidable, choose the minimal substring invariant across the refactors you can foresee — rung 3 shortened a boundary marker from assigned = schools.allocate( to assigned = schools so it matched both sides of its own migration; two other form-brittle checks broke on legitimate reformatting and had to be replaced mid-tranche. Never pin line numbers.
  3. Mutation-prove it can fail, before writing it down. Break the guarded thing, watch the test/check/probe go red, restore. For equality tests, keep a permanent companion mutation test in the suite (rung 3's determinism test ships with a reversed-allocation backend that must always fail the comparison). docs_verify --audit catches vacuous checks; nothing catches a vacuous test but this rule.
  4. Compare typed outcomes, and exclude wall-clock RECURSIVELY. Equality over applied state and event logs, with time-dependent fields scrubbed at every nesting depth — a top-level-only scrub is not enough (commit 863a0fa3). Diagnose flakes to the exact field; never widen an exclusion on a guess.
  5. Tolerate absence in old records. Any test or sweep probe reading the typed record must accept every existing committed root, which predates your feature — assert the attribute exists before reading it, and treat absence as valid, never as failure (the sweep's probe rule; the rung-4 reader-before-writer guardrail).

Style discipline for code steps

  • Match the surrounding code's idiom, naming, and comment density.
  • Comments state constraints the code cannot show, never narrate the change ("why this must hold", not "changed X to Y").
  • Test docstrings name the motivating requirement or record ("Implements R3: ..." / "Regression (run-<id>): ...").
  • Never weaken an existing assertion to make a step pass; that is a failed step, not a passed one.

Exit criteria

  • Exactly one more step checked, with pasted proof; tranche dir committed and pushed.
  • OR the step failed / contradicted the tree: recorded, unchecked, reported back for re-planning.
  • Return to the orchestrator either way.

© AHepi, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .claude/skills/dr-execute-step of AHepi/DeepReason.

Open the folder on GitHubat commit 9607fba

Compare with similar skills

Dr Execute Step next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Dr Execute Step compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Dr Execute Step this skillAHepi/DeepReason141—~1.9kAutomated safety check: PassMIT
Executealirezarezvani/claude-skills28k—~831Automated safety check: PassMIT
Ulw Executecode-yeongyu/oh-my-openagent70k—~6.3kAutomated safety check: PassCustom licence
Executebrycewang-stanford/Auto-Empirical-Research-Skills4.5k—~370Automated safety check: NotesCustom licence
Executionalsk1992/CloddsBot2.9k—~1.7kAutomated safety check: PassMIT
Executive Mentoralirezarezvani/claude-skills28k1 repos~1.8kAutomated safety check: PassMIT

Similar skills

  • Execute

    alirezarezvani/claude-skills

    /cs:execute <decision — Generate a 90-day execution plan with weekly milestones, DRIs, and check-in cadence from an approved decision.

    28k GitHub stars~831 tokensUpdated 1 mo ago
    Product & Project ManagementAuto-check passed
  • Ulw Execute

    code-yeongyu/oh-my-openagent

    Executes a written ulw-plan work plan with Boulder state, evidence ledger, worktree discipline, and parallel subagents.

    70k GitHub stars~6.3k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Execute

    brycewang-stanford/Auto-Empirical-Research-Skills

    Executes all registered notebooks, strips noisy cell metadata, and syncs Jupytext pairs.

    4.5k GitHub stars~370 tokensUpdated 2 days ago
    Data & AnalyticsAuto-check: notes
  • Execution

    alsk1992/CloddsBot

    Execute trades on prediction markets with slippage protection and order management

    2.9k GitHub stars~1.7k tokensUpdated 5 days ago
    Sales & SupportAuto-check passed
  • Executive Mentor

    alirezarezvani/claude-skills

    Adversarial thinking partner for founders and executives. An agent skill from alirezarezvani/claude-skills.

    28k GitHub starsUsed in 1 repo~1.8k tokens
    DevOps & CloudAuto-check passed
  • Executing Plans Inline

    obra/superpowers

    Has the agent carry out an implementation plan itself, task by task in the current session, keeping a ledger, proving each step with a test and ending with one whole-branch review.

    296k GitHub starsUsed in 2 repos~5.1k tokens
    Agent WorkflowsAuto-check passed

More from AHepi/DeepReason

All 30 skills in this repo
  • Pinker Clarity Workflow

    AHepi/DeepReason

    Orchestrate a Steven Pinker-grounded workflow for teaching, explanatory writing, or material that must do both.

    141 GitHub stars~1.5k tokensUpdated 26 days ago
    Auto-check passed
  • Design, deliver, or audit explanations and lessons with a Pinker-informed focus on phenomena, the curse of knowledge, concrete models, active reasoning, feedback, and revision.

    141 GitHub stars~1.8k tokensUpdated 26 days ago
    Auto-check passed
  • Pinker Write For Readers

    AHepi/DeepReason

    Draft, revise, teach, or audit expository prose using Pinker's cognitive approach to style: classic presentation, reader modeling, curse-of-knowledge repair, coherent information order, deliberate…

    141 GitHub stars~2.1k tokensUpdated 26 days ago
    Auto-check passed
  • Example Battery

    AHepi/DeepReason

    Build a battery of concrete instances BEFORE writing or evaluating any definition, pin, or semantic clause (Reed step 1).

    141 GitHub stars~811 tokensUpdated 26 days ago
    Auto-check passed
  • Authoring Skills

    AHepi/DeepReason

    Rules for writing, editing, and retiring skill and workflow files for LLM agents.

    141 GitHub stars~1.7k tokensUpdated 26 days ago
    Auto-check passed
  • Deepreason Orchestrator

    AHepi/DeepReason

    Entry point for any DeepReason problem. An agent skill from AHepi/DeepReason.

    141 GitHub stars~1.1k tokensUpdated 26 days ago
    Auto-check passed

Questions about Dr Execute Step

What does Dr Execute Step do?

Execute exactly one unchecked step from CHECKLIST.md, prove its done-criterion, record the output, and stop. Dr Execute Step is an agent skill from AHepi/DeepReason.md, prove its done-criterion, record the output, and stop.

How do I install Dr Execute Step in Claude Code?

Run `npx skills add AHepi/DeepReason --skill dr-execute-step -a claude-code`. Or copy the skill folder (.claude/skills/dr-execute-step in AHepi/DeepReason) into .claude/skills/dr-execute-step in your project. Claude Code loads it when a task matches its description.

How do I install Dr Execute Step in Codex?

Run `npx skills add AHepi/DeepReason --skill dr-execute-step -a codex`. Or copy the skill folder (.claude/skills/dr-execute-step in AHepi/DeepReason) into .agents/skills/dr-execute-step in your project. Codex loads it when a task matches its description.

Can I use Dr Execute Step in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add AHepi/DeepReason --skill dr-execute-step -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/dr-execute-step, .gemini/skills/dr-execute-step, .github/skills/dr-execute-step and .opencode/skills/dr-execute-step in your project.

What does Dr Execute Step need to run?

Going by SKILL.md and its folder, Dr Execute Step needs the command-line tools its instructions call (git and python). Our summary lists: Python 3.

Does Dr Execute Step access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Dr Execute Step safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Dr Execute Step use?

Dr Execute Step is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Dr Execute Step use?

About 1.9k tokens (SKILL.md is roughly 7.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Dr Execute Step?

Skills that share tags, products or a category with Dr Execute Step: Execute (alirezarezvani/claude-skills, 28k stars), Ulw Execute (code-yeongyu/oh-my-openagent, 70k stars), Execute (brycewang-stanford/Auto-Empirical-Research-Skills, 4.5k stars) and Execution (alsk1992/CloddsBot, 2.9k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Dr Execute Step?

AHepi (a GitHub user) maintains it in AHepi/DeepReason, which has 141 GitHub stars. The repository holds 30 skills in this directory. The repository was last updated on September 10, 2026.

Source: AHepi/DeepReason on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.