Perform a task with a worker agent, critique it with an adversary agent, and apply the accepted critiques with a reconciler agent.
EnglishSettle an unnamed unease about a piece of code by naming the design defect behind it or ruling the unease unfounded, then handing off to the skill that fixes it.
EnglishReplace jargon and imported metaphors in code and prose with concrete, domain-fitting terms a reader understands without translation.
EnglishDiagnose a failing, cancelled, or hanging CI run, explaining why it fails in CI when the same command passes locally.
EnglishDistill a project's transcripts into proposals for eliminating repetitive work: skills, scripts, hooks, rules, or fixes to whatever causes the repetition.
EnglishDo an editorial review of a post or article and report findings ordered by importance.
EnglishFind interface boundaries that leak their implementation and close the leaks, so each module hides one coherent decision behind an interface stated in its callers' terms.
EnglishAttempt a task in a clean-room Claude Code session that has none of this session's context and none of the project's agent instructions, and report the friction log it kept.
EnglishMake a set of things (code, prose, config, docs) consistent along a dimension the user names, by picking a canonical form and conforming every member to it.
EnglishRemove illegal states from data structures by redesigning their types so the illegal states cannot be represented at all.
EnglishUpdate the project's CLAUDE.md to match its current structure and conventions.
EnglishMove code that ended up in the wrong place to where it belongs: the right file, module, layer, or package.
EnglishRename a concept everywhere: propose candidate names with tradeoffs, let the user pick, then sweep the winner through identifiers, prose, docs, and skills.
EnglishRefactor code so each function, class, and module sits at a single level of abstraction and reads top-down as a narrative.
EnglishExtract a DDD-style ubiquitous language glossary.md file from the current conversation and codebase, flagging ambiguities and proposing canonical terms.
English
AGENT SKILLS ATLASIncludes Adversarialize, Concretize, Diagnose CI, and other workflows.
Claude Code
Codex
Cursor
GitHub Copilot
Gemini CLISupported by the Atlas install prompt. Skill-specific requirements vary.
Pack stats
What’s included
Adversarialize
Perform a task with a worker agent, critique it with an adversary agent, and apply the accepted critiques with a reconciler agent.
Allay
Settle an unnamed unease about a piece of code by naming the design defect behind it or ruling the unease unfounded, then handing off to the skill that fixes it.
Concretize
Replace jargon and imported metaphors in code and prose with concrete, domain-fitting terms a reader understands without translation.
Diagnose CI
Diagnose a failing, cancelled, or hanging CI run, explaining why it fails in CI when the same command passes locally.
Distill Transcripts
Distill a project's transcripts into proposals for eliminating repetitive work: skills, scripts, hooks, rules, or fixes to whatever causes the repetition.
Editorialize
Do an editorial review of a post or article and report findings ordered by importance.
Encapsulate
Find interface boundaries that leak their implementation and close the leaks, so each module hides one coherent decision behind an interface stated in its callers' terms.
Friction Log
Attempt a task in a clean-room Claude Code session that has none of this session's context and none of the project's agent instructions, and report the friction log it kept.
Git Split
Split the current git commit into multiple focused, self-contained, easy to
Homogenize
Make a set of things (code, prose, config, docs) consistent along a dimension the user names, by picking a canonical form and conforming every member to it.
Jj Resolve Conflicts
Resolve conflicts in the current jj commit.
Outlaw States
Remove illegal states from data structures by redesigning their types so the illegal states cannot be represented at all.
Reconcile Claude Markdown
Update the project's CLAUDE.md to match its current structure and conventions.
Reconcile Skills
Update a project's SKILL.md files to match the code they describe.
Relocate
Move code that ended up in the wrong place to where it belongs: the right file, module, layer, or package.
Rename Concept
Rename a concept everywhere: propose candidate names with tradeoffs, let the user pick, then sweep the winner through identifiers, prose, docs, and skills.
Stratify
Refactor code so each function, class, and module sits at a single level of abstraction and reads top-down as a narrative.
Ubiquitize Language
Extract a DDD-style ubiquitous language glossary.md file from the current conversation and codebase, flagging ambiguities and proposing canonical terms.
Jj Describe
Write a description for the current jj commit.
Quick start
- Choose your skills. Select individual skills above, or choose the entire pack.
- Copy the install prompt. Use Copy install prompt above the list.
- Send it to your agent. Paste into your coding agent’s chat. Your agent reviews the source and guides the setup.
Try it with your agent
After installation, start with a request like:
Use the Adversarialize skill to help with my task.
Ask me for the context you need before getting started.Requirements & license
- A coding agent that can access your project files and install skills.
- Your agent checks each skill’s instructions, dependencies, and permissions during setup.
- Reported license for the skills shown: MIT. The author’s terms apply.