@iceinvein/agent Skills

GitHub Actions

Install agent skills into AI coding tools

Add to Cursor Add to VS Code Run on PlugMCP

1{2  "mcpServers": {3    "iceinvein-agent-skills": {4      "command": "npx",5      "args": [6        "-y",7        "@iceinvein/[email protected]"8      ]9    }10  }11}

.cursor/mcp.jsonMCP server

Transport: stdio
Status: Local

Power your agent

Add this server

The publisher's own command. Not a PlugMCP endpoint.

Add to Cursor Add to VS Code Run on PlugMCP

Host a copy with our keys and logs.

Local command

npx -y @iceinvein/[email protected]

Claude Code has no install link for a local command. Copy the command below and run it in a terminal.

shClaude Code

1claude mcp add iceinvein-agent-skills -- npx -y @iceinvein/agent-skills@0.23.9

Run in a terminal.

jsonClaude Desktop

1{2  "mcpServers": {3    "iceinvein-agent-skills": {4      "command": "npx",5      "args": [6        "-y",7        "@iceinvein/[email protected]"8      ]9    }10  }11}

claude_desktop_config.json

Local · npmPublisher installStep-by-step: add this in Claude

Remote URL or a command on your machine? How the two installs differ

Run on PlugMCP

Host a copy with our keys and logs.

A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.

Launch

Server facts

Local

Transport: stdio
Version: 0.23.9
Last checked: 2026-10-28

Stars: —
Tools: —
Uptime: —

Tools

No tool list yet. Remote servers get one after the first health check.

FAQ

Questions

How do I install @iceinvein/agent Skills?

Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.

Does @iceinvein/agent Skills need an API key?

No required environment variable is listed on this record.

What transport does @iceinvein/agent Skills use?

@iceinvein/agent Skills runs locally over stdio.

When was @iceinvein/agent Skills last checked?

Last checked 2026-10-28. Install agent skills into AI coding tools

From the publisher

Documentation

@iceinvein/agent-skills

Thinking frameworks from foundational software engineering texts, installed as skills for AI coding agents.

Each skill encodes a specific design methodology (Brooks, Parnas, Rams, Feathers, etc.) so your agent doesn't just write code; it reasons about structure, coupling, complexity, and perception the way these authors taught us to.

Install

Requires Bun. Install it with curl -fsSL https://bun.sh/install | bash.

bunx @iceinvein/agent-skills install <skill> [<skill> ...]   # one or more
bunx @iceinvein/agent-skills install                         # no args: opens picker

Auto-detects Claude Code, Cursor, Codex, or Gemini CLI. Use --tool claude to target one, -g to install globally.

Claude Code plugin

Each skill is also a plugin in the iceinvein-skills marketplace:

/plugin marketplace add iceinvein/agent-skills
/plugin install <skill>@iceinvein-skills

Plugin skills are namespaced: invoke them as /<skill>:<skill> (e.g. /sluice:sluice).
Third-party marketplaces do not auto-update by default. Update from /plugin.
Activation modes do not apply: sluice's hooks always run while the plugin is enabled, and terse is never auto-activated.
Do not install sluice through both routes, or its hooks run twice. Its statusline segment is bunx-only.
magpie, migrate and scope-estimate need Bun on PATH.
improve-my-codebase's plugin carries every audit, so under the plugin route its audits do not need installing separately.

Skills

Migration

Skill	What it does
migrate	Walks a legacy codebase through probe, enumerate, seam, extract, parity, queue, adjudicate and handoff: measured surface coverage, an empirically derived capability seam, cited requirements, a parity plan, a batch decision queue, and work items emitted to markdown, GitHub or a flow target with coverage and forecast read back. Bundles a Bun CLI that enforces the coverage arithmetic and phase ordering instead of trusting it. Claude Code only.
scope-estimate	The lighter path for a proposal: read-only discovery, a capability register with status and disposition, sized functional requirements, contracts-first work packages, and a build schedule across engineers x worktrees at fast, planning and floor rates. Bundles a Bun scheduler that refuses a plan missing or double-counting a requirement, so every figure comes from one reproducible run. Claude Code only.

Skill

What it does

migrate

Walks a legacy codebase through probe, enumerate, seam, extract, parity, queue, adjudicate and handoff: measured surface coverage, an empirically derived capability seam, cited requirements, a parity plan, a batch decision queue, and work items emitted to markdown, GitHub or a flow target with coverage and forecast read back. Bundles a Bun CLI that enforces the coverage arithmetic and phase ordering instead of trusting it. Claude Code only.

scope-estimate

The lighter path for a proposal: read-only discovery, a capability register with status and disposition, sized functional requirements, contracts-first work packages, and a build schedule across engineers x worktrees at fast, planning and floor rates. Bundles a Bun scheduler that refuses a plan missing or double-counting a requirement, so every figure comes from one reproducible run. Claude Code only.

Code Architecture

Skills that analyze how code is structured: module boundaries, coupling, complexity, contracts, and evolution over time.

Skill	Based on	What it does
complexity-accountant	Ousterhout, A Philosophy of Software Design	Treats complexity as a finite budget. Every abstraction must justify itself by being deep (simple interface, rich functionality), not shallow.
contract-enforcer	Meyer, Object-Oriented Software Construction	Before writing non-trivial functions, articulates preconditions, postconditions, invariants, and failure contracts, then verifies the implementation satisfies them.
module-secret-auditor	Parnas, On the Criteria for Decomposing Systems into Modules	Every module should hide exactly one design decision that's likely to change. Boundaries drawn by change-reason, not by noun or technical layer.
coupling-auditor	Constantine & Yourdon, Structured Design	Classifies coupling between modules on the spectrum (data > stamp > control > common > content) and recommends one step down.
simplicity-razor	Hickey, Simple Made Easy	Distinguishes simple (not interleaved) from easy (familiar). Names the strands, detects complecting, and decomplects or justifies.
error-strategist	Duffy & Abrahams	Classifies errors (bug/recoverable/fatal), assigns exception safety guarantees (nothrow/strong/basic), and designs recovery boundaries.
seam-finder	Feathers, Working Effectively with Legacy Code	Before changing existing code, finds seams: places where behavior can be altered without editing the code at that point. Minimal incision, preserve the unknown.
evolution-analyzer	Lehman, Laws of Software Evolution	Evaluates changes through the Laws of Software Evolution: assessing trajectory, debt impact, and whether the system is adapting or accreting.
cohesion-analyzer	Constantine & Yourdon, Structured Design	Classifies module focus on the 7-level cohesion spectrum (coincidental → functional), identifies mixed responsibilities, and proposes specific split lines.
demeter-enforcer	Lieberherr, Law of Demeter	Detects chain violations, parameter drilling, and hidden traversal. Fixes with "tell, don't ask" or parameter narrowing to reduce structural coupling.
dependency-direction-auditor	Martin, Clean Architecture	Traces imports across architectural layers, classifies direction violations by severity (hard/soft/transitive), and recommends specific inversions.
type-driven-designer	Wlaschin, Domain Modeling Made Functional; Minsky, "Make Illegal States Unrepresentable"	Uses the type system to encode business rules: branded types for validated values, discriminated unions for state machines, elimination of primitive obsession.
cqs-auditor	Meyer, Object-Oriented Software Construction	Classifies functions as commands (change state) or queries (return data), detects mixed violations, and separates side effects from return values.

UI & Visual Design

Skills that evaluate interfaces through perception, cognition, and design principles.

Skill	Based on	What it does
rams-design-audit	Dieter Rams, Ten Principles of Good Design	Every visual element must earn its presence. If removing it loses nothing, remove it. Less but better.
cognitive-load-auditor	Jeff Johnson, Designing with the Mind in Mind	Evaluates UI against Miller's Law (working memory), Hick's Law (decision time), Fitts's Law (target size), and cognitive load theory.
gestalt-reviewer	Gestalt psychology	Checks that visual grouping matches logical grouping through proximity, similarity, closure, continuity, and figure-ground analysis.

System Design & Integration

Skills for reviewing high-level architecture, messaging patterns, and data flow.

Skill	Based on	What it does
design-review	Brooks, The Design of Design	Interactive interview that tests conceptual integrity, constraint exploitation, removal discipline, and scope control.
codebase-architecture	–	Architecture review for existing codebases or structured design for new projects. Includes a comprehensive patterns reference.
integration-pattern-auditor	Hohpe & Woolf, Enterprise Integration Patterns	Names the integration pattern (channel, router, transformer, endpoint), verifies delivery guarantees, and identifies missing infrastructure like dead letter queues and idempotency.
unidirectional-flow-enforcer	Elm Architecture	Enforces unidirectional state flow in UI applications: events up, state down, mutations in one place. Detects bidirectional mutations and cascading effects.
event-design-reviewer	Evans, Vernon, Dahan (DDD)	Events should describe what happened in the domain, not what changed in the database. Applies the domain expert test, evaluates payload design, and catches CRUD naming smells.
bounded-context-auditor	Evans, Domain-Driven Design	Detects linguistic fractures (polysemous terms, contested models), draws context maps with relationship types, and identifies leaking language and shared model pollution.
port-adapter-auditor	Cockburn, Hexagonal Architecture	Maps driving and driven ports, classifies boundary health (clean/missing/leaking/wrong-language), and ensures core testability and infrastructure swappability.
idempotency-guardian	Helland, Idempotence Is Not a Medical Condition	Audits mutation endpoints and event handlers for retry safety: classifies natural idempotency, checks protection mechanisms, and evaluates side effect safety.
composability-auditor	McIlroy, Pike, Thompson (Unix Philosophy)	Identifies reuse blockers (hidden I/O, non-standard interfaces, statefulness) and decomposes self-sufficient units into composable pieces.
temporal-coupling-detector	Beck, Fowler (software engineering canon)	Detects hidden ordering dependencies (two-phase init, method order, invisible preconditions) and enforces them through types, parameters, and factory patterns.

Verification

Skills that check code does what it was meant to do, with evidence rather than opinion.

Skill	Based on	What it does
bug-hunter	QA scenario testing; intent-first reproduction	Proves logic bugs with failing repros checked against stated intent. Writes scenarios before reading the code, and reports bugs, ambiguities, and covered scenarios. Claude Code only.

Communication

Skills that control how the agent communicates: output compression, token efficiency, and response style.

Skill	Based on	What it does
terse	Caveman (refined)	Professional output compression: cuts ~20-30% of tokens by eliminating LLM verbosity habits (preambles, hedging, trailing summaries) while keeping proper grammar. Three levels: clean, tight, sharp.

Career & Writing

Skills for producing job application materials that read like a human wrote them, match the role, and align with the resume.

Skill	Role	What it does
cover-letter	Suite router	Dispatches to write, audit, rewrite, or persona subcommands. Shares conventions for file I/O (PDF, DOCX, MD/MDX, URL), output formats, and writing principles across the suite.
cover-letter-write	Generator	Generates a cover letter from resume + job description. Outputs markdown, DOCX, and PDF. Enforces human prose (burstiness, concrete verbs, no filler openers), maps every claim to a resume bullet, and covers top JD requirements without keyword stuffing.
cover-letter-audit	Scorer	Scores a letter on 100 points across content/fit, structure, voice/humanness, and correctness. Detects AI-generated-content signals (burstiness, 50+ AI phrases including sentimental cliches, TTR, em-dashes, passive voice) and produces prioritized fixes.
cover-letter-rewrite	Reviser	Audit-driven targeted rewrite of an existing letter. Focus modes: humanize, align, tighten, structure, tone. Preserves voice where it already works.
cover-letter-persona	Voice profile	Reusable writing personas using the NNGroup 4-dimension tone framework (funny-serious, formal-casual, respectful-irreverent, enthusiastic-matter-of-fact) adapted for professional correspondence.

Workflow

Skills that shape how work gets done rather than analyzing code.

Skill What it does

sluice Routes work by change shape into four channels (bypass, fast, main, deep) and applies only the rules each channel needs, so a one-line fix does not pay the cost of a multi-subsystem build. Eight rules live as one-liners in the router; the full treatment sits in references read only on friction. Each run closes with a ledger read out of the session transcript: elapsed, tools, tokens, and what every dispatched agent cost. Claude Code only. Conflicts with the superpowers plugin, which requires its own fixed pipeline up front for anything that adds to or changes what the software does, not just code edits; disable superpowers before installing.

Skill	What it does
sluice	Routes work by change shape into four channels (`bypass`, `fast`, `main`, `deep`) and applies only the rules each channel needs, so a one-line fix does not pay the cost of a multi-subsystem build. Eight rules live as one-liners in the router; the full treatment sits in references read only on friction. Each run closes with a ledger read out of the session transcript: elapsed, tools, tokens, and what every dispatched agent cost. Claude Code only. Conflicts with the superpowers plugin, which requires its own fixed pipeline up front for anything that adds to or changes what the software does, not just code edits; disable superpowers before installing.

Orchestration

Skills that compose the other audit skills into higher-level workflows.

Skill What it does

improve-my-codebase Runs every applicable audit skill in parallel against your codebase, ranks findings by convergence (multiple audits agreeing) and per-file rollup, and writes a prioritized improvement report. Default sweeps everything; positional args switch modes (quick, diff, interactive) or narrow scope (focus <area>, module <path>).

magpie Interactive PR review pipeline. Runs five parallel specialist subagents (security, bugs, performance, code-smells, architecture), dedupes findings, applies a critic rubric, peer-reviews via codex exec, and serves an interactive HTML report for selecting findings to post via gh. A local mode reviews the current branch diff before a PR exists and hands the findings back in the terminal. Claude Code only.

Skill	What it does
improve-my-codebase	Runs every applicable audit skill in parallel against your codebase, ranks findings by convergence (multiple audits agreeing) and per-file rollup, and writes a prioritized improvement report. Default sweeps everything; positional args switch modes (`quick`, `diff`, `interactive`) or narrow scope (`focus <area>`, `module <path>`).
magpie	Interactive PR review pipeline. Runs five parallel specialist subagents (security, bugs, performance, code-smells, architecture), dedupes findings, applies a critic rubric, peer-reviews via `codex exec`, and serves an interactive HTML report for selecting findings to post via `gh`. A local mode reviews the current branch diff before a PR exists and hands the findings back in the terminal. Claude Code only.

Commands

bunx @iceinvein/agent-skills install <skill> [<skill> ...] [--tool <tool>] [--activation <mode>] [-g]
bunx @iceinvein/agent-skills install                                                          # picker
bunx @iceinvein/agent-skills browse                                                           # alias
bunx @iceinvein/agent-skills remove <skill> [-g]
bunx @iceinvein/agent-skills update <skill> [-g]
bunx @iceinvein/agent-skills list
bunx @iceinvein/agent-skills info <skill>

install accepts multiple skill names and installs each in turn. Run install (or browse) with no args to open an interactive picker: arrow keys to navigate, space to toggle, enter to install all selected, Esc to cancel.

Flag
`--tool <tool>`	Target a specific tool: `claude`, `cursor`, `codex`, `gemini`
`--activation <mode>`	For skills that support it: `session` (manual `/skill`) or `global` (auto via `SessionStart` hook). Claude Code only.
`-g, --global`	Install to home directory (available in all projects)

Updating Skills

Each skill is versioned independently (see skill.json inside each skill). The CLI tracks what you installed in .agent-skills.lock and can pull newer versions from GitHub.

bunx @iceinvein/agent-skills update <skill>     # update one skill
bunx @iceinvein/agent-skills update --all       # update everything in the lockfile
bunx @iceinvein/agent-skills list               # show installed versions

update re-fetches the latest manifest and files from this repo, removes the old install, and re-installs cleanly. Your chosen activation mode (session vs. global) is preserved across updates, as is the set of tools you installed for. Use -g to update globally-installed skills.

After any command other than update, the CLI runs a background freshness check (at most once every 24 hours) and prints a one-line notice if anything is out of date:

⚠ Updates available: some-skill (v0.2.0 → v0.3.0)
  Run: agent-skills update --all

The check is best-effort: it silently skips on network failure and only hits GitHub once per day per project. To force a re-check sooner, delete the lastUpdateCheck field in .agent-skills.lock (or just run any command after 24 hours have elapsed).

To pin to the current version, do nothing: skills are not auto-updated. update is opt-in.

Activation Modes (Claude Code)

Some skills (like terse and sluice) support activation modes. Pick one at install time:

session: invoke the skill manually with /<skill> each session
global: auto-activate every Claude Code session via a SessionStart hook in .claude/settings.json

Each skill declares its own default: terse defaults to session, sluice defaults to global because a router that is not always on does not route.

If a skill declares activation support and you're installing for Claude Code interactively, the CLI prompts you. Use --activation session or --activation global for scripted installs. remove strips the hook cleanly; update preserves your choice across version bumps.

Supported Tools

Claude Code, Cursor, Codex and Gemini CLI. Most skills install into all four; bug-hunter, magpie, migrate, scope-estimate and sluice are Claude Code only, and the CLI skips the other tools for them.

How It Works

The CLI fetches skills from GitHub and installs them into the right locations for your tool:

Tool	Install path
Claude Code	`.claude/skills/` + `.claude/settings.json`
Cursor	`.cursor/rules/` + `.cursor/mcp.json`
Codex	Appended to `AGENTS.md`
Gemini CLI	`.gemini/skills/` + `.gemini/settings.json`

Skills install to the current project by default. Use -g to install to your home directory so the skill is available everywhere. A .agent-skills.lock file tracks installations for update and remove.

Contributing

Releasing: how versions are bumped, tagged, and published to npm

License

MIT

@iceinvein/agent Skills

GitHub Actions

Install agent skills into AI coding tools

Devops Developer Tools

Add to Cursor Add to VS Code Run on PlugMCP

GitHub npm Website

Add to Cursor Add to VS Code Run on PlugMCP

1{2  "mcpServers": {3    "iceinvein-agent-skills": {4      "command": "npx",5      "args": [6        "-y",7        "@iceinvein/[email protected]"8      ]9    }10  }11}

.cursor/mcp.jsonMCP server

Transport: stdio
Status: Local

Power your agent

Add this server

The publisher's own command. Not a PlugMCP endpoint.

Add to Cursor Add to VS Code Run on PlugMCP

Host a copy with our keys and logs.

Local command

npx -y @iceinvein/[email protected]

Claude Code has no install link for a local command. Copy the command below and run it in a terminal.

shClaude Code

1claude mcp add iceinvein-agent-skills -- npx -y @iceinvein/agent-skills@0.23.9

Run in a terminal.

jsonClaude Desktop

1{2  "mcpServers": {3    "iceinvein-agent-skills": {4      "command": "npx",5      "args": [6        "-y",7        "@iceinvein/[email protected]"8      ]9    }10  }11}

claude_desktop_config.json

Local · npmPublisher installStep-by-step: add this in Claude

Remote URL or a command on your machine? How the two installs differ

Run on PlugMCP

Host a copy with our keys and logs.

A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.

Launch

Server facts

Local

Transport: stdio
Version: 0.23.9
Last checked: 2026-10-28

Stars: —
Tools: —
Uptime: —

Tools

No tool list yet. Remote servers get one after the first health check.

FAQ

Questions

How do I install @iceinvein/agent Skills?

Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.

Does @iceinvein/agent Skills need an API key?

No required environment variable is listed on this record.

What transport does @iceinvein/agent Skills use?

@iceinvein/agent Skills runs locally over stdio.

When was @iceinvein/agent Skills last checked?

Last checked 2026-10-28. Install agent skills into AI coding tools

From the publisher

Documentation

@iceinvein/agent-skills

Thinking frameworks from foundational software engineering texts, installed as skills for AI coding agents.

Install

Requires Bun. Install it with curl -fsSL https://bun.sh/install | bash.

bunx @iceinvein/agent-skills install <skill> [<skill> ...]   # one or more
bunx @iceinvein/agent-skills install                         # no args: opens picker

Auto-detects Claude Code, Cursor, Codex, or Gemini CLI. Use --tool claude to target one, -g to install globally.

Claude Code plugin

Each skill is also a plugin in the iceinvein-skills marketplace:

/plugin marketplace add iceinvein/agent-skills
/plugin install <skill>@iceinvein-skills

Plugin skills are namespaced: invoke them as /<skill>:<skill> (e.g. /sluice:sluice).
Third-party marketplaces do not auto-update by default. Update from /plugin.
Activation modes do not apply: sluice's hooks always run while the plugin is enabled, and terse is never auto-activated.
Do not install sluice through both routes, or its hooks run twice. Its statusline segment is bunx-only.
magpie, migrate and scope-estimate need Bun on PATH.
improve-my-codebase's plugin carries every audit, so under the plugin route its audits do not need installing separately.

Skills

Migration

Skill	What it does
migrate	Walks a legacy codebase through probe, enumerate, seam, extract, parity, queue, adjudicate and handoff: measured surface coverage, an empirically derived capability seam, cited requirements, a parity plan, a batch decision queue, and work items emitted to markdown, GitHub or a flow target with coverage and forecast read back. Bundles a Bun CLI that enforces the coverage arithmetic and phase ordering instead of trusting it. Claude Code only.
scope-estimate	The lighter path for a proposal: read-only discovery, a capability register with status and disposition, sized functional requirements, contracts-first work packages, and a build schedule across engineers x worktrees at fast, planning and floor rates. Bundles a Bun scheduler that refuses a plan missing or double-counting a requirement, so every figure comes from one reproducible run. Claude Code only.

Skill

What it does

migrate

scope-estimate

Code Architecture

Skills that analyze how code is structured: module boundaries, coupling, complexity, contracts, and evolution over time.

Skill	Based on	What it does
complexity-accountant	Ousterhout, A Philosophy of Software Design	Treats complexity as a finite budget. Every abstraction must justify itself by being deep (simple interface, rich functionality), not shallow.
contract-enforcer	Meyer, Object-Oriented Software Construction	Before writing non-trivial functions, articulates preconditions, postconditions, invariants, and failure contracts, then verifies the implementation satisfies them.
module-secret-auditor	Parnas, On the Criteria for Decomposing Systems into Modules	Every module should hide exactly one design decision that's likely to change. Boundaries drawn by change-reason, not by noun or technical layer.
coupling-auditor	Constantine & Yourdon, Structured Design	Classifies coupling between modules on the spectrum (data > stamp > control > common > content) and recommends one step down.
simplicity-razor	Hickey, Simple Made Easy	Distinguishes simple (not interleaved) from easy (familiar). Names the strands, detects complecting, and decomplects or justifies.
error-strategist	Duffy & Abrahams	Classifies errors (bug/recoverable/fatal), assigns exception safety guarantees (nothrow/strong/basic), and designs recovery boundaries.
seam-finder	Feathers, Working Effectively with Legacy Code	Before changing existing code, finds seams: places where behavior can be altered without editing the code at that point. Minimal incision, preserve the unknown.
evolution-analyzer	Lehman, Laws of Software Evolution	Evaluates changes through the Laws of Software Evolution: assessing trajectory, debt impact, and whether the system is adapting or accreting.
cohesion-analyzer	Constantine & Yourdon, Structured Design	Classifies module focus on the 7-level cohesion spectrum (coincidental → functional), identifies mixed responsibilities, and proposes specific split lines.
demeter-enforcer	Lieberherr, Law of Demeter	Detects chain violations, parameter drilling, and hidden traversal. Fixes with "tell, don't ask" or parameter narrowing to reduce structural coupling.
dependency-direction-auditor	Martin, Clean Architecture	Traces imports across architectural layers, classifies direction violations by severity (hard/soft/transitive), and recommends specific inversions.
type-driven-designer	Wlaschin, Domain Modeling Made Functional; Minsky, "Make Illegal States Unrepresentable"	Uses the type system to encode business rules: branded types for validated values, discriminated unions for state machines, elimination of primitive obsession.
cqs-auditor	Meyer, Object-Oriented Software Construction	Classifies functions as commands (change state) or queries (return data), detects mixed violations, and separates side effects from return values.

UI & Visual Design

Skills that evaluate interfaces through perception, cognition, and design principles.

Skill	Based on	What it does
rams-design-audit	Dieter Rams, Ten Principles of Good Design	Every visual element must earn its presence. If removing it loses nothing, remove it. Less but better.
cognitive-load-auditor	Jeff Johnson, Designing with the Mind in Mind	Evaluates UI against Miller's Law (working memory), Hick's Law (decision time), Fitts's Law (target size), and cognitive load theory.
gestalt-reviewer	Gestalt psychology	Checks that visual grouping matches logical grouping through proximity, similarity, closure, continuity, and figure-ground analysis.

System Design & Integration

Skills for reviewing high-level architecture, messaging patterns, and data flow.

Skill	Based on	What it does
design-review	Brooks, The Design of Design	Interactive interview that tests conceptual integrity, constraint exploitation, removal discipline, and scope control.
codebase-architecture	–	Architecture review for existing codebases or structured design for new projects. Includes a comprehensive patterns reference.
integration-pattern-auditor	Hohpe & Woolf, Enterprise Integration Patterns	Names the integration pattern (channel, router, transformer, endpoint), verifies delivery guarantees, and identifies missing infrastructure like dead letter queues and idempotency.
unidirectional-flow-enforcer	Elm Architecture	Enforces unidirectional state flow in UI applications: events up, state down, mutations in one place. Detects bidirectional mutations and cascading effects.
event-design-reviewer	Evans, Vernon, Dahan (DDD)	Events should describe what happened in the domain, not what changed in the database. Applies the domain expert test, evaluates payload design, and catches CRUD naming smells.
bounded-context-auditor	Evans, Domain-Driven Design	Detects linguistic fractures (polysemous terms, contested models), draws context maps with relationship types, and identifies leaking language and shared model pollution.
port-adapter-auditor	Cockburn, Hexagonal Architecture	Maps driving and driven ports, classifies boundary health (clean/missing/leaking/wrong-language), and ensures core testability and infrastructure swappability.
idempotency-guardian	Helland, Idempotence Is Not a Medical Condition	Audits mutation endpoints and event handlers for retry safety: classifies natural idempotency, checks protection mechanisms, and evaluates side effect safety.
composability-auditor	McIlroy, Pike, Thompson (Unix Philosophy)	Identifies reuse blockers (hidden I/O, non-standard interfaces, statefulness) and decomposes self-sufficient units into composable pieces.
temporal-coupling-detector	Beck, Fowler (software engineering canon)	Detects hidden ordering dependencies (two-phase init, method order, invisible preconditions) and enforces them through types, parameters, and factory patterns.

Verification

Skills that check code does what it was meant to do, with evidence rather than opinion.

Skill	Based on	What it does
bug-hunter	QA scenario testing; intent-first reproduction	Proves logic bugs with failing repros checked against stated intent. Writes scenarios before reading the code, and reports bugs, ambiguities, and covered scenarios. Claude Code only.

Communication

Skills that control how the agent communicates: output compression, token efficiency, and response style.

Skill	Based on	What it does
terse	Caveman (refined)	Professional output compression: cuts ~20-30% of tokens by eliminating LLM verbosity habits (preambles, hedging, trailing summaries) while keeping proper grammar. Three levels: clean, tight, sharp.

Career & Writing

Skills for producing job application materials that read like a human wrote them, match the role, and align with the resume.

Skill	Role	What it does
cover-letter	Suite router	Dispatches to write, audit, rewrite, or persona subcommands. Shares conventions for file I/O (PDF, DOCX, MD/MDX, URL), output formats, and writing principles across the suite.
cover-letter-write	Generator	Generates a cover letter from resume + job description. Outputs markdown, DOCX, and PDF. Enforces human prose (burstiness, concrete verbs, no filler openers), maps every claim to a resume bullet, and covers top JD requirements without keyword stuffing.
cover-letter-audit	Scorer	Scores a letter on 100 points across content/fit, structure, voice/humanness, and correctness. Detects AI-generated-content signals (burstiness, 50+ AI phrases including sentimental cliches, TTR, em-dashes, passive voice) and produces prioritized fixes.
cover-letter-rewrite	Reviser	Audit-driven targeted rewrite of an existing letter. Focus modes: humanize, align, tighten, structure, tone. Preserves voice where it already works.
cover-letter-persona	Voice profile	Reusable writing personas using the NNGroup 4-dimension tone framework (funny-serious, formal-casual, respectful-irreverent, enthusiastic-matter-of-fact) adapted for professional correspondence.

Workflow

Skills that shape how work gets done rather than analyzing code.

Skill What it does

Skill	What it does
sluice	Routes work by change shape into four channels (`bypass`, `fast`, `main`, `deep`) and applies only the rules each channel needs, so a one-line fix does not pay the cost of a multi-subsystem build. Eight rules live as one-liners in the router; the full treatment sits in references read only on friction. Each run closes with a ledger read out of the session transcript: elapsed, tools, tokens, and what every dispatched agent cost. Claude Code only. Conflicts with the superpowers plugin, which requires its own fixed pipeline up front for anything that adds to or changes what the software does, not just code edits; disable superpowers before installing.

Orchestration

Skills that compose the other audit skills into higher-level workflows.

Skill What it does

Skill	What it does
improve-my-codebase	Runs every applicable audit skill in parallel against your codebase, ranks findings by convergence (multiple audits agreeing) and per-file rollup, and writes a prioritized improvement report. Default sweeps everything; positional args switch modes (`quick`, `diff`, `interactive`) or narrow scope (`focus <area>`, `module <path>`).
magpie	Interactive PR review pipeline. Runs five parallel specialist subagents (security, bugs, performance, code-smells, architecture), dedupes findings, applies a critic rubric, peer-reviews via `codex exec`, and serves an interactive HTML report for selecting findings to post via `gh`. A local mode reviews the current branch diff before a PR exists and hands the findings back in the terminal. Claude Code only.

Commands

bunx @iceinvein/agent-skills install <skill> [<skill> ...] [--tool <tool>] [--activation <mode>] [-g]
bunx @iceinvein/agent-skills install                                                          # picker
bunx @iceinvein/agent-skills browse                                                           # alias
bunx @iceinvein/agent-skills remove <skill> [-g]
bunx @iceinvein/agent-skills update <skill> [-g]
bunx @iceinvein/agent-skills list
bunx @iceinvein/agent-skills info <skill>

Flag
`--tool <tool>`	Target a specific tool: `claude`, `cursor`, `codex`, `gemini`
`--activation <mode>`	For skills that support it: `session` (manual `/skill`) or `global` (auto via `SessionStart` hook). Claude Code only.
`-g, --global`	Install to home directory (available in all projects)

Updating Skills

Each skill is versioned independently (see skill.json inside each skill). The CLI tracks what you installed in .agent-skills.lock and can pull newer versions from GitHub.

bunx @iceinvein/agent-skills update <skill>     # update one skill
bunx @iceinvein/agent-skills update --all       # update everything in the lockfile
bunx @iceinvein/agent-skills list               # show installed versions

After any command other than update, the CLI runs a background freshness check (at most once every 24 hours) and prints a one-line notice if anything is out of date:

⚠ Updates available: some-skill (v0.2.0 → v0.3.0)
  Run: agent-skills update --all

To pin to the current version, do nothing: skills are not auto-updated. update is opt-in.

Activation Modes (Claude Code)

Some skills (like terse and sluice) support activation modes. Pick one at install time:

session: invoke the skill manually with /<skill> each session
global: auto-activate every Claude Code session via a SessionStart hook in .claude/settings.json

Each skill declares its own default: terse defaults to session, sluice defaults to global because a router that is not always on does not route.

Supported Tools

Claude Code, Cursor, Codex and Gemini CLI. Most skills install into all four; bug-hunter, magpie, migrate, scope-estimate and sluice are Claude Code only, and the CLI skips the other tools for them.

How It Works

The CLI fetches skills from GitHub and installs them into the right locations for your tool:

Tool	Install path
Claude Code	`.claude/skills/` + `.claude/settings.json`
Cursor	`.cursor/rules/` + `.cursor/mcp.json`
Codex	Appended to `AGENTS.md`
Gemini CLI	`.gemini/skills/` + `.gemini/settings.json`

Contributing

Releasing: how versions are bumped, tagged, and published to npm

License

MIT