The Jury
tech-leads-club/agent-skills
A skill your agent uses when a question, decision, plan, tradeoff, or claim needs a rigorous verdict and one perspective is not enough.
A skill your agent uses when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a…
Install Adversarial Review
$ npx skills add DanMcInerney/architect-loop --skill adversarial-review -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install DanMcInerney/architect-loop adversarial-review --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/DanMcInerney/architect-loop.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/adversarial-review .claude/skills/adversarial-review && 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 "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .claude/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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/DanMcInerney/architect-loop/tree/main/skills/adversarial-reviewType 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 DanMcInerney/architect-loop --skill adversarial-review -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install DanMcInerney/architect-loop adversarial-review --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/adversarial-review .agents/skills/adversarial-review && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .agents/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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 DanMcInerney/architect-loop --skill adversarial-review -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install DanMcInerney/architect-loop adversarial-review --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/adversarial-review .cursor/skills/adversarial-review && 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 "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .cursor/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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/DanMcInerney/architect-loop.git --path skills/adversarial-review--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 DanMcInerney/architect-loop --skill adversarial-review -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install DanMcInerney/architect-loop adversarial-review --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/adversarial-review .gemini/skills/adversarial-review && 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 "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .gemini/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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 DanMcInerney/architect-loop adversarial-reviewInstalls 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 DanMcInerney/architect-loop --skill adversarial-review -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/adversarial-review .github/skills/adversarial-review && 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 "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .github/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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 DanMcInerney/architect-loop --skill adversarial-review -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install DanMcInerney/architect-loop adversarial-review --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/adversarial-review .opencode/skills/adversarial-review && 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 "adversarial-review" agent skill from https://github.com/DanMcInerney/architect-loop/tree/main/skills/adversarial-review into .opencode/skills/adversarial-review/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "adversarial-review", 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
adversarial-review- GitHub stars
- 626
- Token cost
- ~1.1k tokens
- SKILL.md length
- 568 words
- Files
- 1
- Skills in repo
- 11
- Repo updated
- First seen
- Licence
- MIT
At a glance
A skill your agent uses when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a…
- Works in 3 steps: spec attack (post-/to-spec) → revise and decompose → decomposition stress test (pre-freeze)
- The architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence
- SKILL.md covers Calibration, Stage 1: spec attack…, Stage 2: revise and decompose and Stage 3: decomposition stress…, plus 3 more sections
- Calls git
What it does
Adversarial Review is an agent skill from DanMcInerney/architect-loop. Use when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a revised spec, decompose it (issues plus checks), and stress-test its own decomposition before returning all drafts. Never commits, touches the tracker, or edits product code.
Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Agent Workflows, covering Load testing and Subagents. The repository describes itself as: Super optimized /goal loop. Massive token savings and higher quality. Smart model designs and reviews, cheaper model builds.. The licence is MIT.
When your agent uses it
- The architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence
- Fold the surviving findings into a revised spec
- Decompose it (issues plus checks)
- Stress-test its own decomposition before returning all drafts
Example prompts
- “/adversarial-review”
Workflow steps
3 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit 28dca7d. 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
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
Adversarial Review loads about 1.1k tokens when it runs. Until then it costs about 93 tokens; SKILL.md has 568 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 DanMcInerney/architect-loop at commit 28dca7d, republished under its MIT licence (© DanMcInerney). 568 words, ~1,070 tokens.
Download SKILL.mdSave it as
.claude/skills/adversarial-review/SKILL.md (or your agent's skills folder).- name
- adversarial-review
- description
- Use when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a revised spec, decompose it (issues plus checks), and stress-test its own decomposition before returning all drafts. Never commits, touches the tracker, or edits product code.
- effort
- high
Adversarial Review
You are a fresh strategist with no stake in this plan surviving. Break it
first; rebuild only from what the attack proves. Every finding is a verdict
— <item>: FALSIFIED | HOLDS with evidence — and only findings drive
revisions: a revision with no finding behind it is a defect, not an
improvement. Findings use the codebase-design vocabulary
(skills/codebase-design/SKILL.md) — a plan that drifts from it is itself
a finding.
Calibration
Flag only gaps that affect correctness, the stated requirements, or documented project invariants — no stylistic preferences.
Stage 1: spec attack (post-/to-spec)
Attack the draft spec on its own terms:
- Contradictions between sections — goal vs non-goals, target flow vs design constraints, assumption vs validation strategy.
- Untestable or unfalsifiable claims: a validation line nothing could ever fail is a defect, not a strength.
- Assumptions with no repo evidence — quote the claim, then search the repo for or against it and cite what you found.
- Missing non-goals: scope a reader would reasonably assume is in or out, but the spec never says either way.
- Scope beyond the stated goal — a requirement the goal doesn't license.
Every finding quotes the claim or cites the spec file:line.
Stage 2: revise and decompose
Only after the attack is exhausted: fold each FALSIFIED finding's fix into
a revised spec on to-spec's exact template — evidence-backed revisions
only — and record what survived unevidenced under ## Assumptions. Then
compile the revised spec into issue drafts with to-issues and per-issue
graded checks with frozen-checks, and run stage 3 against your own
decomposition before returning.
Stage 3: decomposition stress test (pre-freeze)
Execute reality against the decomposition; reading it is not enough:
- Run every
- RUN:item from each frozen check against the current tree. A mechanical check with no->expectation is itself a check defect. - Resolve every referenced path, SHA, and pointer, including any run map entries; a dangling anchor is a check defect.
- Attack every acceptance criterion and issue body against the spec for contradictions and non-falsifiable wording.
- Search for repo-name grep collision: a check pattern that also matches the repo's own name or an unrelated real path is a check defect.
- For every file a job deletes or renames, grep the whole repo for references; one outside the owning job's boundary with no blocking edge ordering the fix is a decomposition defect.
- Run
git check-ignore <path>on every new artifact path a job will create; an ignored path is a decomposition defect.
Show full SKILL.md (169 more words)Show less
Reporting
Return to the orchestrator: the revised spec path, the issue-draft and
check-draft paths, and a flat findings ledger, one line per item:
<check id, clause, or claim>: FALSIFIED | HOLDS plus the command output
or quoted evidence behind it, so every revision traces to its finding.
Close with any assumptions that survived review unevidenced. Do not
summarize, encourage, or soften a finding.
Boundary
You revise the spec and draft issues and checks; you never commit, never touch the tracker, and never edit product code or the mutable test suite — publishing and the freeze are orchestrator actions after you return. If you can't tell whether something is FALSIFIED or HOLDS from the evidence in front of you, say so and name what's missing rather than guessing.
Vocabulary
Use the factory glossary exactly: module, interface, implementation, seam, adapter, depth, leverage, locality, run, tracking issue, issue, slice, frozen check, check-runner, strategist, builder, orchestrator, factory branch, worktree, job report, verdict, ruling, digest, hard stop. Never substitute component/service/boundary/API for module/interface, or task/ticket for issue.
© DanMcInerney, 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 skills/adversarial-review of DanMcInerney/architect-loop.
Open the folder on GitHubat commit 28dca7d
Compare with similar skills
Adversarial Review 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 |
|---|---|---|---|---|---|---|
| Adversarial Review this skillDanMcInerney/architect-loop | 626 | — | ~1.1k | Automated safety check: Pass | MIT | |
| The Jurytech-leads-club/agent-skills | 7k | — | ~4.5k | Automated safety check: Pass | CC-BY-4.0 | |
| Dialecticumputun/cc-thingz | 485 | — | ~756 | Automated safety check: Notes | MIT | |
| Grillingpietheinstrengholt/rssmonster | 564 | 31 repos | ~510 | Automated safety check: Pass | MIT | |
| Workflow Orchestrationvxcozy/workflow-orchestration | 116 | — | ~1k | Automated safety check: Pass | MIT | |
| Idea Genieboshu2/agentops | 448 | 1 repos | ~1.8k | Automated safety check: Pass | Apache-2.0 |
Similar skills
Dialectic
umputun/cc-thingz
Prove and counter-prove a statement using parallel agents to eliminate confirmation bias.
Grilling
pietheinstrengholt/rssmonster
Grill the user relentlessly about a plan, decision, or idea.
Workflow Orchestration
vxcozy/workflow-orchestration
Disciplined task execution with planning, verification, and self-improvement loops.
Idea Genie
boshu2/agentops
Brainstorm evidence-backed options for what to build, or stress-test an idea.
Loop Change Verifier
cobusgreyling/loop-engineering
Acts as a skeptical checker for changes an implementer sub-agent made, running the tests, confirming the diff scope and returning approve, reject or escalate to a human.
More from DanMcInerney/architect-loop
All 11 skills in this repoArchitect Research
DanMcInerney/architect-loop
A skill your agent uses when the user asks for discovery-scale research that informs a decision: brainstorming a project or feature, choosing a technology, or requests like "research X", "what's the…
Architect
DanMcInerney/architect-loop
A skill your agent uses when the user asks to architect, run or continue the autonomous software factory, turn a goal into a hardened tracker issue plan, dispatch builder jobs, grade finished work…
Architect Fast
DanMcInerney/architect-loop
A skill your agent uses when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into…
Final Review
DanMcInerney/architect-loop
A skill your agent uses for the closing whole-run review in the architect factory — the only model review in the loop: dispatched by the orchestrator, at finish, to one fresh strategist subagent…
Frozen Checks
DanMcInerney/architect-loop
A skill your agent uses when the strategist drafts per-issue graded checks after decomposition and before builder dispatch.
TDD
DanMcInerney/architect-loop
Test-driven development for factory builders. An agent skill from DanMcInerney/architect-loop.
Related
Categories
Questions about Adversarial Review
What does Adversarial Review do?
A skill your agent uses when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a…. Adversarial Review is an agent skill from DanMcInerney/architect-loop. Use when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a revised spec, decompose it (issues plus checks), and stress-test its own decomposition before returning all drafts.
When should I use Adversarial Review?
Adversarial Review fits situations like: the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence; fold the surviving findings into a revised spec; decompose it (issues plus checks); stress-test its own decomposition before returning all drafts.
How do I install Adversarial Review in Claude Code?
Run `npx skills add DanMcInerney/architect-loop --skill adversarial-review -a claude-code`. Or copy the skill folder (skills/adversarial-review in DanMcInerney/architect-loop) into .claude/skills/adversarial-review in your project. Claude Code loads it when a task matches its description.
How do I install Adversarial Review in Codex?
Run `npx skills add DanMcInerney/architect-loop --skill adversarial-review -a codex`. Or copy the skill folder (skills/adversarial-review in DanMcInerney/architect-loop) into .agents/skills/adversarial-review in your project. Codex loads it when a task matches its description.
Can I use Adversarial Review 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 DanMcInerney/architect-loop --skill adversarial-review -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/adversarial-review, .gemini/skills/adversarial-review, .github/skills/adversarial-review and .opencode/skills/adversarial-review in your project.
What does Adversarial Review need to run?
Going by SKILL.md and its folder, Adversarial Review needs the command-line tools its instructions call (git).
Does Adversarial Review 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 Adversarial Review 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 Adversarial Review use?
Adversarial Review 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 Adversarial Review use?
About 1.1k tokens (SKILL.md is roughly 4.3k 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 Adversarial Review?
Skills that share tags, products or a category with Adversarial Review: The Jury (tech-leads-club/agent-skills, 7k stars), Dialectic (umputun/cc-thingz, 485 stars), Grilling (pietheinstrengholt/rssmonster, 564 stars) and Workflow Orchestration (vxcozy/workflow-orchestration, 116 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Adversarial Review?
DanMcInerney (a GitHub user) maintains it in DanMcInerney/architect-loop, which has 626 GitHub stars. The repository holds 11 skills in this directory. The repository was last updated on September 13, 2026.
Source: DanMcInerney/architect-loop on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.