Show Me Your Work Decision Log
cursor/plugins
Keeps a TSV decision log for long or unattended agent runs, one row per decision with what, why, evidence and result, so a reviewer can check the work later.
Runs an iterative verification loop on Google Antigravity: make one edit, run a test command, and reschedule until the command passes or the iteration budget runs out.
Install Loop Until Verified
$ npx skills add jtaroreh/agystack --skill loop -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install jtaroreh/agystack loop --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/loop .claude/skills/loop && rm -rf skills-srcUse ~/.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/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .claude/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/jtaroreh/agystack/tree/main/skills/loopType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add jtaroreh/agystack --skill loop -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install jtaroreh/agystack loop --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/loop .agents/skills/loop && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .agents/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add jtaroreh/agystack --skill loop -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install jtaroreh/agystack loop --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/loop .cursor/skills/loop && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .cursor/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/jtaroreh/agystack.git --path skills/loop--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add jtaroreh/agystack --skill loop -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install jtaroreh/agystack loop --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/loop .gemini/skills/loop && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .gemini/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install jtaroreh/agystack loopInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add jtaroreh/agystack --skill loop -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/loop .github/skills/loop && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .github/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add jtaroreh/agystack --skill loop -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install jtaroreh/agystack loop --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/jtaroreh/agystack.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/loop .opencode/skills/loop && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "loop" agent skill from https://github.com/jtaroreh/agystack/tree/main/skills/loop into .opencode/skills/loop/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "loop", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
loop- GitHub stars
- 103
- Token cost
- ~607 tokens
- SKILL.md length
- 273 words
- Files
- 2 (incl. references)
- Skills in repo
- 6
- Repo updated
- First seen
- Licence
- MIT
At a glance
Runs an iterative verification loop on Google Antigravity: make one edit, run a test command, and reschedule until the command passes or the iteration budget runs out.
- Works in 5 steps: Frame. Read target files. Generate… → Step. Apply one discrete edit or… → Verify. Run the verification command.… → …
- Fixing code until a specific test command passes
- SKILL.md covers How It Works, Invocation, State Machine and Arguments
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
You start it with /loop, a goal and a verify command that exits 0 on success, with optional limits on iterations and the delay between them, defaulting to 10 attempts and 30 seconds. State is kept in a state.json file under the conversation's scratch folder, and the native Antigravity schedule tool provides the delays so the agent does not need to poll.
Each round follows a state machine. It frames the task by reading target files and running the verify command for a baseline, applies one edit or hypothesis, runs the verification and captures output and exit code, then decides. A pass marks the state converged and finishes. A failure with attempts left records the change, reverts regressions when hillclimbing, sets a timer and ends the turn without calling more tools. Running out of attempts marks it exhausted and hands back a summary table. Open-ended tasks without a single exit code are sent to /goal or the autonomous run playbook.
When your agent uses it
- Fixing code until a specific test command passes
- Hillclimbing a measurable result with one change per iteration
- Testing a hypothesis against a verification command with a bounded budget
- Avoiding manual polling while waiting between attempts
Example prompts
- “/loop make the parser tests pass --verify "npm test -- parser" --max-iterations 8”
- “Loop until green: fix the failing lint errors, verifying with npm run lint.”
- “Keep trying fixes for the flaky checkout test, running pytest tests/test_checkout.py each time.”
- “Run the loop with a 60 second interval and a budget of 5 iterations.”
Requirements
- Google Antigravity with its schedule tool
- A verification command that exits 0 on success
Workflow steps
5 steps, taken from the first numbered list in SKILL.md.
- Frame. Read target files. Generate baseline evidence by executing the verify command via run_command or bash. Record the state in…
- Step. Apply one discrete edit or hypothesis.
- Verify. Run the verification command. Capture stdout, stderr, and exit code.
- Decide.
- Wakeup. On schedule trigger, the agent is reactively woken up, reads state.json, and executes the next step.
What it can do on your machine
Read from SKILL.md and the folder at commit d1a0466. 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
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
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
Loop Until Verified loads about 607 tokens when it runs, and up to ~878 if it reads all its reference files. Until then it costs about 53 tokens; SKILL.md has 273 words of instructions outside code blocks.
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 jtaroreh/agystack at commit d1a0466, republished under its MIT licence (© jtaroreh). 273 words, ~607 tokens.
Download SKILL.mdSave it as
.claude/skills/loop/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- loop
- description
- Run iterative verification hillclimbs against a concrete test command using Antigravity reactive scheduling. Use for /loop, 'loop until green', or automated hypothesis testing with a verification predicate.
Loop
Run iterative verification loops and hillclimbs on Antigravity without manual polling.
How It Works
The /loop skill drives discrete edits against a verification command (--verify "<command>"). It maintains iteration state in <appDataDir>/brain/<conversation-id>/scratch/loop/state.json and uses the native Antigravity schedule tool to manage reactive iteration delays.
For open-ended autonomous tasks where no single bash exit code determines completion, use the /goal slash command or the Autonomous run playbook instead. For standing background jobs or periodic status checks, use the schedule tool directly.
Invocation
Format:
/loop <goal> --verify "<command>" [--max-iterations N] [--interval S]
Default values:
max-iterations: 10interval: 30 seconds
State Machine
- Frame. Read target files. Generate baseline evidence by executing the verify command via
run_commandor bash. Record the state in<appDataDir>/brain/<conversation-id>/scratch/loop/state.json. - Step. Apply one discrete edit or hypothesis.
- Verify. Run the verification command. Capture stdout, stderr, and exit code.
- Decide.
- If verify passes: log success, mark state
CONVERGED, and finish. - If verify fails and iterations remain: record the delta, revert regressions if hillclimbing, and set a reactive timer using
schedule(DurationSeconds=interval, Prompt="Loop iteration wakeup for <goal>", TimerCondition="never"). - IMPORTANT: Calling
schedulereturns immediately and does not pause execution. You MUST stop calling tools immediately after scheduling to end your turn and let the timer fire. - If budget is reached: mark
EXHAUSTEDand hand back to the user with a summary table.
- If verify passes: log success, mark state
- Wakeup. On schedule trigger, the agent is reactively woken up, reads
state.json, and executes the next step.
Arguments
| Argument | Description | Default |
|---|---|---|
goal | Description of the desired state | Required |
--verify | Command that exits 0 on success | Required |
--max-iterations | Maximum number of attempts | 10 |
--interval | Delay between iterations in seconds | 30 |
© jtaroreh, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file (references) in skills/loop of jtaroreh/agystack.
- SKILL.md
- references/loop-protocol.md
Open the folder on GitHubat commit d1a0466
Compare with similar skills
Loop Until Verified 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.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Loop Until Verified this skilljtaroreh/agystack | 103 | — | ~607 | Automated safety check: Pass | MIT | |
| Show Me Your Work Decision Logcursor/plugins | 10k | 9 repos | ~1.6k | Automated safety check: Pass | None | |
| PUA Looptanweai/pua | 20k | — | ~1.1k | Automated safety check: Pass | MIT | |
| Flowguard Task Guardmajiayu000/spellbook | 286 | — | ~2.1k | Automated safety check: Pass | MIT | |
| ULW Loopcode-yeongyu/oh-my-openagent | 70k | — | ~3.2k | Automated safety check: Pass | Custom licence | |
| Superloopy Evidence Loopbeefiker/superloopy | 111 | 1 repos | ~5.9k | Automated safety check: Pass | MIT |
Similar skills
- Official
PUA Loop
tanweai/pua
Runs an unattended iterate-until-verified loop in which a user-set verify command, not the agent's own claim, decides when the task is finished.
Flowguard Task Guard
majiayu000/spellbook
Single entry point that routes long or ambiguous agent tasks, checks live state, bounds autonomous loops and leaves a resumable handoff.
ULW Loop
code-yeongyu/oh-my-openagent
Runs a long task as a checkpointed goal loop: it creates goals, mirrors each step into a todo list, gathers evidence per criterion and lands every goal before the next.
Superloopy Evidence Loop
beefiker/superloopy
Runs a light task loop where each goal criterion passes only when a real evidence artifact exists, with progress stored in a `.superloopy` folder.
Goal Objective Drafter
QwenLM/qwen-code
Turns a vague intention into a /goal objective with one outcome, numbered yes-or-no checks, guardrails, a budget and a block protocol, without starting the work.
More from jtaroreh/agystack
agystack Runtime Setup
jtaroreh/agystack
Configures agystack's model tiers per role and its execution runtime, choosing between local subagents and Cloud Run jobs for large parallel swarms.
Deslop
jtaroreh/agystack
Clean AI slop and defensiveness out of code and diffs before commit.
agystack Engineering Framework
jtaroreh/agystack
agystack's agentic engineering framework and rigor mode. Runs poteto-mode playbooks, isolated subagents, and cloud swarms. Use for /agystack, agystack, or…
Principle Attack The Premise
jtaroreh/agystack
Apply when handed a problem, constraint, or design you are tempted to accept as given.
Principle Test Behavior Not Implementation
jtaroreh/agystack
Apply when writing or reviewing tests. An agent skill from jtaroreh/agystack.
Related
Categories
Questions about Loop Until Verified
What does Loop Until Verified do?
Runs an iterative verification loop on Google Antigravity: make one edit, run a test command, and reschedule until the command passes or the iteration budget runs out. You start it with /loop, a goal and a verify command that exits 0 on success, with optional limits on iterations and the delay between them, defaulting to 10 attempts and 30 seconds.json file under the conversation's scratch folder, and the native Antigravity schedule tool provides the delays so the agent does not need to poll.
When should I use Loop Until Verified?
Loop Until Verified fits situations like: fixing code until a specific test command passes; hillclimbing a measurable result with one change per iteration; testing a hypothesis against a verification command with a bounded budget; avoiding manual polling while waiting between attempts.
How do I install Loop Until Verified in Claude Code?
Run `npx skills add jtaroreh/agystack --skill loop -a claude-code`. Or copy the skill folder (skills/loop in jtaroreh/agystack) into .claude/skills/loop in your project. Claude Code loads it when a task matches its description.
How do I install Loop Until Verified in Codex?
Run `npx skills add jtaroreh/agystack --skill loop -a codex`. Or copy the skill folder (skills/loop in jtaroreh/agystack) into .agents/skills/loop in your project. Codex loads it when a task matches its description.
Can I use Loop Until Verified 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 jtaroreh/agystack --skill loop -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/loop, .gemini/skills/loop, .github/skills/loop and .opencode/skills/loop in your project.
What does Loop Until Verified need to run?
SKILL.md names no scripts, command-line tools or credentials: Loop Until Verified is instructions for the agent only. Our summary lists: Google Antigravity with its schedule tool; A verification command that exits 0 on success.
Does Loop Until Verified access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Loop Until Verified 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 Loop Until Verified use?
Loop Until Verified 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 Loop Until Verified use?
About 607 tokens (SKILL.md is roughly 2.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 271 tokens, read only when the agent opens those files.
What are the alternatives to Loop Until Verified?
Skills that share tags, products or a category with Loop Until Verified: Show Me Your Work Decision Log (cursor/plugins, 10k stars), PUA Loop (tanweai/pua, 20k stars), Flowguard Task Guard (majiayu000/spellbook, 286 stars) and ULW Loop (code-yeongyu/oh-my-openagent, 70k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Loop Until Verified?
jtaroreh (a GitHub user) maintains it in jtaroreh/agystack, which has 103 GitHub stars. The repository holds 6 skills in this directory. The repository was last updated on October 1, 2026.
Source: jtaroreh/agystack on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.