~ / rigs / stevencarpenter / agents

stevencarpenter/agents

General-purpose AI agent definitions, reusable skills, and registry tooling

↗ GitHub ★ 1 MIT updated 21d ago collection Claude CodeCodex
stack: Python
share on X
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
~4.9k tokens
Moderate · median rig: 2.3k · breakdown
GUARDRAILS
0/5
No committed guardrails · details
Quality 5/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit stevencarpenter/agents/agents ./rig-agents/agents
$ npx degit stevencarpenter/agents/skills ./rig-agents/skills

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ npx degit stevencarpenter/agents/skills/blind-review .claude/skills/blind-review
$ npx degit stevencarpenter/agents/skills/dast-staging-guidelines .claude/skills/dast-staging-guidelines
$ npx degit stevencarpenter/agents/skills/data-engineering-guidelines .claude/skills/data-engineering-guidelines
$ npx degit stevencarpenter/agents/skills/diagramming-guidelines .claude/skills/diagramming-guidelines
$ npx degit stevencarpenter/agents/skills/java-guidelines .claude/skills/java-guidelines
$ npx degit stevencarpenter/agents/skills/jj-guidelines .claude/skills/jj-guidelines
$ npx degit stevencarpenter/agents/skills/kotlin-guidelines .claude/skills/kotlin-guidelines
$ npx degit stevencarpenter/agents/skills/python-guidelines .claude/skills/python-guidelines
$ npx degit stevencarpenter/agents/skills/rust-guidelines .claude/skills/rust-guidelines
$ npx degit stevencarpenter/agents/skills/sast-triage-guidelines .claude/skills/sast-triage-guidelines
$ npx degit stevencarpenter/agents/skills/scala-guidelines .claude/skills/scala-guidelines
$ npx degit stevencarpenter/agents/skills/security-review-guidelines .claude/skills/security-review-guidelines
$ npx degit stevencarpenter/agents/skills/spark-guidelines .claude/skills/spark-guidelines
$ npx degit stevencarpenter/agents/skills/spark-pyspark-guidelines .claude/skills/spark-pyspark-guidelines
$ npx degit stevencarpenter/agents/skills/spark-scala-guidelines .claude/skills/spark-scala-guidelines
$ npx degit stevencarpenter/agents/skills/sql-guidelines .claude/skills/sql-guidelines
$ npx degit stevencarpenter/agents/skills/supply-chain-guidelines .claude/skills/supply-chain-guidelines
$ npx degit stevencarpenter/agents/skills/swift-guidelines .claude/skills/swift-guidelines
$ npx degit stevencarpenter/agents/skills/technical-writing-guidelines .claude/skills/technical-writing-guidelines
$ npx degit stevencarpenter/agents/skills/terraform-guidelines .claude/skills/terraform-guidelines
$ npx degit stevencarpenter/agents/skills/tool-priority .claude/skills/tool-priority
$ npx degit stevencarpenter/agents/skills/tutoring-guidelines .claude/skills/tutoring-guidelines
$ npx degit stevencarpenter/agents/skills/typescript-guidelines .claude/skills/typescript-guidelines
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/blind-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/cli-tool-designer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/code-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/data-engineer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>-specialist/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>-auditor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/docs-lookup/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>-writer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.<redacted>-tutor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/grafana-read/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/hippo-query/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/java-implementer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/java-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/jj-workflow/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/kotlin-implementer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/kotlin-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/linear-ops/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/mcp-server-builder/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/pr-manager/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/python-implementer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/python-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/railway-read/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-api-designer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-async-concurrency-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-idiom-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-implementer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-performance-profiler/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-test-strategist/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-toolchain-auditor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/rust-unsafe-ffi-auditor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/sast-scanner/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/scala-implementer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/scala-reviewer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/secrets-auditor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/security-auditor/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/slide-designer/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/spark-performance-profiler/agent.md
$ curl -fsSL --create-dirs -o .claude/agents/agent.md https://raw.githubusercontent.com/stevencarpenter/agents/main/agents/spark-pyspark-implementer/agent.md

This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:

{
  "permissions": {
    "deny": [
      "Read(./.env)",
      "Read(./.env.*)",
      "Read(**/.env)",
      "Read(~/.ssh/**)",
      "Bash(git push --force:*)",
      "Read(**/*.pem)",
      "Read(.env)",
      "Read(~/.aws/**)",
      "Read(**/*.key)",
      "Read(**/.env.*)",
      "Bash(git push --force*)",
      "Bash(rm -rf:*)",
      "Bash(rm -rf *)",
      "Bash(sudo *)",
      "Bash(sudo:*)",
      "Bash(rm -rf /)",
      "Bash(git push -f:*)",
      "Read(.env.*)",
      "Bash(git reset --hard*)",
      "Bash(rm -rf /*)"
    ],
    "ask": [
      "Bash(git push:*)",
      "Bash(git push *)",
      "Bash(gh pr merge *)",
      "Bash(git commit:*)",
      "Bash(rm *)",
      "Bash(gh pr merge:*)",
      "Bash(gh pr create *)",
      "Bash(git reset --hard:*)",
      "Bash(npm publish:*)",
      "Bash(git rebase *)"
    ]
  }
}

Skills (23)

Subagents (55)

Explore
model: haiku
Use when locating code by pattern, grepping for symbols or keywords, or answering "where is X defined" / "which files reference Y" — a fast read-only search agent, not a review or analysis agent.
architecture-reviewer
model: inherit
Use when evaluating system design decisions, module boundaries, API shapes, data model choices, or the overall structure of a feature or codebase before significant implementation begins.
blind-reviewer
model: inherit
Use when performing a context-blind review of code with comments, docstrings, and documentation stripped. PoC/experimental — dispatched by the blind-review skill against a stripped scratch workspace,
cli-tool-designer
model: inherit
Use when designing or implementing a CLI tool — including argument parsing, subcommand structure, output formatting, error handling, and cross-platform behavior.
code-reviewer
model: inherit
Use when reviewing a diff that spans multiple languages, or non-code changes (config, scripts, CI, docs), or when no single-language reviewer fits. For language-specific depth prefer rust-idiom-review
data-engineer
model: inherit
Use when designing or building data-engineering pipelines on a lakehouse — Delta Lake or Apache Iceberg tables, the medallion model, catalog governance (Unity Catalog) — realtime/streaming ingestion (
debugging-specialist
model: inherit
Use when diagnosing a bug, test failure, crash, hang, or unexpected behavior — especially when the root cause is not obvious from the error message alone.
dependency-auditor
model: inherit
Use when auditing third-party dependencies for known vulnerabilities, typosquat or ownership risk, unpinned images, and lockfile hygiene across pip, cargo, npm, and Go ecosystems.
docs-lookup
model: haiku
Use when fetching current library, framework, or API documentation for a named package or symbol — resolving a library ID and pulling its docs. Not for deciding which library to use or synthesizing an
documentation-writer
model: inherit
Use when writing or improving technical documentation — README files, API references, architecture decision records, runbooks, or inline doc comments. For tutorials, how-to guides, and conceptual expl
engineering-tutor
model: inherit
Use when the user wants to learn by doing — coached through a ticket, bug, or design with questions and hints while they write all the code themselves. Socratic follow-ups, TDD shaping, no generated i
grafana-read
model: haiku
Use when running an already-specified Prometheus/Loki/Pyroscope query, or looking up a dashboard, datasource, incident, alert, or on-call schedule by name/UID — retrieval only. Not for composing a Pro
hippo-query
model: haiku
Use when retrieving information from the Hippo local knowledge base — past shell activity, Claude sessions, browser history, memory documents, or entities — for a query that already states what to loo
java-implementer
model: inherit
Use when implementing Java code that must be idiomatic, immutable where possible, well-typed, and aligned with the repo's build and framework conventions.
java-reviewer
model: inherit
Use when reviewing Java code for null safety, immutability, generics correctness, exception handling, thread safety, or API design.
jj-workflow
model: inherit
Use when working with Jujutsu (jj) in git-colocated repos — confirm jj state, apply jj-guidelines, and use local just jj-* helpers when present.
kotlin-implementer
model: inherit
Use when implementing Kotlin code that must be null-safe, immutable, coroutine-correct, and aligned with the repo's build and platform conventions.
kotlin-reviewer
model: inherit
Use when reviewing Kotlin code for null safety, coroutine/structured-concurrency correctness, immutability, ADT modeling, or scope-function misuse.
linear-ops
model: haiku
Use when creating or updating Linear issues, comments, labels, or projects from a fully-specified request — title, team, and any field values are already known. Not for triaging ambiguous backlog prio
mcp-server-builder
model: inherit
Use when building, extending, or debugging an MCP (Model Context Protocol) server — including tool definitions, resource handlers, transport setup, and schema design.
pr-manager
model: inherit
Use when managing pull request lifecycle — writing PR descriptions, triaging review feedback, coordinating merge readiness, or deciding between PR strategies (single vs split, squash vs merge).
python-implementer
model: inherit
Use when implementing Python code that must be typed, tested, and idiomatic — including scripts, libraries, CLI tools, and data pipelines.
python-reviewer
model: inherit
Use when reviewing Python code for type safety, correctness, idiomatic patterns, error handling, test coverage, or subprocess/filesystem risks.
railway-read
model: haiku
Use when checking Railway project/service status, logs, metrics, deployments, domains, variables, or feature flags — read-only lookups with a named project/service already in hand. Never for deploys,
rust-api-designer
model: inherit
Use when designing or reviewing Rust public APIs, crate boundaries, error models, feature flags, builders, or semver-sensitive changes.
rust-async-concurrency-reviewer
model: inherit
Use when reviewing Rust async, Tokio, concurrency, scheduling, cancellation, backpressure, channels, locks, or task lifetime behavior.
rust-idiom-reviewer
model: inherit
Use when reviewing Rust code for idiomatic API design, ownership, error handling, async correctness, unsafe usage, performance traps, or maintainability. For deep Tokio, channel, or task-lifetime revi
rust-implementer
model: inherit
Use when implementing Rust code that must be idiomatic, minimal, test-backed, and aligned with existing crate boundaries.
rust-performance-profiler
model: inherit
Use when investigating Rust performance, allocation, cloning, lock contention, async throughput, hot paths, benchmarks, or profiling evidence.
rust-test-strategist
model: inherit
Use when reviewing or designing Rust test strategy — unit/integration/doc test layout, property-based tests, fuzz targets, coverage gaps, flaky async tests, or test maintainability. For general code r
rust-toolchain-auditor
model: inherit
Use when auditing Rust Cargo.toml files, workspace layout, dependency hygiene, feature flags, MSRV, supply-chain policy, build profiles, or CI gates. For .rs code review prefer rust-idiom-reviewer; fo
rust-unsafe-ffi-auditor
model: inherit
Use when auditing Rust unsafe code, FFI, raw pointers, lifetimes, Send or Sync impls, memory layout, aliasing, or soundness boundaries.
sast-scanner
model: inherit
Use when running static security analysis over a repo — Semgrep sweeps, Bandit, gosec, cargo audit — and triaging raw scanner output into verified true positives.
scala-implementer
model: inherit
Use when implementing Scala code that must be idiomatic, immutable, type-safe, and aligned with the repo's effect system and build conventions.
scala-reviewer
model: inherit
Use when reviewing Scala code for immutability, ADT modeling, effect handling, implicit/given hygiene, collection performance, or type clarity.
secrets-auditor
model: inherit
Use when scanning a repo's working tree and git history for committed credentials, tokens, private keys, and env-file hygiene problems.
security-auditor
model: inherit
Use when threat-modeling an application, reviewing a diff or whole repo for exploitable weaknesses, or coordinating a security assessment across static, dependency, and staging specialists.
slide-designer
model: inherit
Use when designing or building a technical presentation — structuring the narrative, writing action/assertion titles, choosing the right chart for a message, or generating decks in Marp, reveal.js, Sl
spark-performance-profiler
model: inherit
Use when investigating Apache Spark 4 pipeline performance — explain plans, AQE/skew joins, shuffle volume, partition pruning, small-files problems, Python UDF bottlenecks, caching misuse, and executo
spark-pyspark-implementer
model: inherit
Use when implementing Apache Spark 4 data pipelines in Python — batch ETL, lakehouse MERGE/CDC, typed PySpark transforms, pandas UDFs, and Structured Streaming jobs with transformWithStateInPandas.

Similar rigs

copied ✓