Development Claude Skills
8 673 development skills from 759 authors. They write and change code. Half of them fit into 1 830 tokens or less — that is what one costs your context window when the agent loads it. 1 213 ship runnable scripts rather than instructions alone. 42 of them cannot work without an MCP server, most often rube. We also found 1 239 copies of these same skills sitting in other people's repositories — counted once here, not 1 239 times.
8 673 unique 759 authors 10 updated this month 1 368 from vendors
Systematic external reference searching across documentation, open-source repositories, and web resources. Use when working with unfamiliar libraries, APIs, or when needing production-quality implementation examples.
Checks whether a new version of a repository preserves the behavior observed by tests on the old version. Use this skill when comparing two versions of code to detect regressions, verify refactoring safety, validate bug fixes don't break existing functionality, or ensure backward compatibility. Detects differences in function outputs, exceptions, observable states, and performance between versions. Generates reports highlighting potential regressions (critical, high, medium, low severity), improvements, and areas requiring verification. Triggers when users ask to check for regressions between versions, compare test behavior across versions, verify behavior preservation, or validate that changes don't break existing tests.
Locate root causes of failing regression tests by analyzing code changes, error messages, and test dependencies. Use when regression tests start failing after code changes, investigating test failures in CI/CD, debugging flaky tests, or understanding why previously passing tests now fail. Analyzes git diffs, stack traces, test output, and dependency changes to produce structured markdown reports ranking likely causes. Triggers when users ask to find why tests are failing, debug regression failures, investigate test breakage, or analyze failing test suites.
Instruments programs to record execution information for deterministic replay debugging. Use when debugging hard-to-reproduce bugs (race conditions, timing issues, intermittent failures, heisenbugs), reproducing production failures, or analyzing complex execution sequences. Records non-deterministic events (I/O, threading, randomness, time) to enable exact replay of program executions. Supports Python, JavaScript, Java, and C/C++ with both custom instrumentation and existing replay tools.
Instruments programs to capture execution traces specifically for reproducing reported bugs, enabling consistent replay and diagnosis of failures. Use this skill when you need to reproduce a bug, capture execution traces for debugging, instrument code to record program behavior, generate replay scripts for bug reproduction, diagnose hard-to-reproduce failures, or perform deterministic replay of program execution. Triggers when users ask to instrument code for tracing, capture execution traces, reproduce bugs, generate replay scripts, or enable deterministic debugging.
Compares old and new requirement documents, analyzes code repository impact, and generates detailed modification plans. Use when Claude needs to: (1) Compare requirement versions and identify changes, (2) Map requirement changes to code components, (3) Identify components to modify, delete, or add, (4) Analyze dependencies and integration points, (5) Assess test impact, (6) Generate comprehensive modification plans in Markdown format. Supports text/Markdown requirements and analyzes feature-level, functional, and API changes.
Explains runtime errors and compilation failures with actionable debugging guidance. Use when Python or Java code throws runtime exceptions (NullPointerException, TypeError, AttributeError, etc.), compilation errors (syntax errors, type mismatches, import failures), or dependency issues. Analyzes error messages, stack traces, and code context to identify root causes and provide concrete fixes with examples. Distinct from test-related errors - focuses on errors during normal code execution and build processes.
Instruments authentication, authorization, and input-handling code paths to monitor security-relevant events and states at runtime. Use this skill when developers need to add security monitoring and logging to their code, including tracking authentication attempts (login/logout), authorization decisions (access control checks), input validation failures, session management events, and other security-critical operations. Supports Python, JavaScript/TypeScript, and Java with structured logging patterns. Triggers when users ask to add security instrumentation, monitor security events, log authentication/authorization, track security-sensitive operations, or add security observability to their codebase.
Detect semantic-level bugs by analyzing whether code behavior matches its intended purpose inferred from function/variable names, comments, docstrings, and documentation. Use when users need to: (1) Find logic errors where implementation contradicts stated intent, (2) Identify off-by-one errors and boundary mismatches, (3) Detect inverted logic or wrong operators, (4) Catch missing edge case handling, (5) Verify code matches its documentation. Highlights mismatches between intent and implementation across multiple programming languages.
Analyzes and verifies semantic equivalence between two functions, classes, or modules by examining control flow, data flow, and observable behavior. Use when comparing code implementations (refactored vs original, different implementations of same functionality, migration verification), determining if two code artifacts produce identical behavior, identifying behavioral differences between code versions, or validating that code changes preserve semantics. Supports formal reasoning and symbolic execution approaches.
Analyzes a repository and its test suite to generate customized mutation operators tailored to the project. Use this skill when setting up mutation testing, improving test quality, or generating project-specific mutants. Considers language constructs, business logic patterns, API calls, data types, and code complexity to produce effective mutation operators that maximize test sensitivity while avoiding trivial or equivalent mutations. Generates prioritized operator sets (high/medium/low value) with impact summaries. Triggers when users ask to generate mutation operators, customize mutation testing, analyze code for mutation opportunities, or create project-specific mutants.
Identify bug-introducing commits using semantic analysis that extends traditional SZZ algorithm. Distinguishes semantic changes from refactorings or code movements using control-flow and data-flow similarity analysis. Use when analyzing bug-fix commits to trace back to bug-introducing changes, investigating software evolution, conducting empirical studies on defect prediction, or reducing false positives in bug localization. Supports git repositories and provides explanations for why commits are identified as bug-introducing.
Automatically extract abstract finite-state models in SMV/NuSMV format from source code (C/C++, Java, Python) for formal model checking. Use when users need to: (1) Generate SMV models from program code for verification, (2) Extract state-transition models from protocol implementations, (3) Analyze control flow and data flow to construct formal models, (4) Create models for checking safety and liveness properties, (5) Convert imperative code to declarative state machines. Particularly effective for protocol implementations, concurrent systems, and control logic with clear state transitions.
Generate implementation code and tests from written specifications. Use when the user provides specifications (natural language descriptions, formal specs, requirements documents, API specs) and asks Claude to implement the described functionality. Supports data structures, algorithms, classes, functions, and includes automatic test generation to validate implementation against specification.
Generate formal specifications including preconditions, postconditions, invariants, and contracts from code or requirements. Use this skill when documenting APIs, creating formal verification annotations, defining function contracts, specifying class invariants, writing design-by-contract code, or preparing code for formal verification. Supports multiple specification languages including JML, ACSL, Dafny, Eiffel contracts, and documentation annotations.
Instrument programs (Python, C/C++, Java) to capture snapshots of key program states at runtime, including variables, memory, and call stacks. Use when you need to debug complex issues, reproduce test cases, prepare traces for formal verification, or analyze program execution. Supports manual instrumentation points, automatic function/method instrumentation, and conditional triggers. Outputs structured JSON snapshots for debugging, replay, and verification workflows.
Automatically migrate Spring MVC applications to Spring Boot. Use when you need to modernize a Spring MVC project to Spring Boot while preserving functionality. The skill analyzes the codebase, updates build configuration (Maven/Gradle), migrates annotations, converts XML configuration to Java/properties, updates controllers and tests, and creates the Spring Boot main application class. Creates git commits for each migration phase and generates a comprehensive summary. Supports both Maven and Gradle projects.
Verify code correctness statically against specifications using type checking, contract verification, and formal methods. Use when: (1) Verifying type safety and null safety in Python, Java, or C/C++ code, (2) Checking design-by-contract specifications (preconditions, postconditions, invariants), (3) Validating code against formal specifications, (4) Ensuring code quality and correctness before runtime, (5) Finding potential bugs through static analysis. Supports Python (mypy, contracts), Java (javac, JML), and provides verification scripts and contract specification guidelines.
> Performs symbolic execution to detect potential errors by exploring execution paths, solving path constraints, and generating test inputs. Use when you need to analyze code for bugs like null dereferences, division by zero, buffer overflows, or assertion violations. Also use to generate test inputs that exercise different code paths, find edge cases, or explore all reachable program states. Supports Python, Java, and C/C++ through manual symbolic execution techniques and integration with tools like KLEE, angr, Z3, and Symbolic PathFinder.
Identifies bug-introducing commits using SZZ-style analysis based on bug-fixing commits, commit history, and code blame information. Use this skill when you need to trace bugs back to their origin, identify which commits introduced bugs, analyze bug-fix commits to find root causes, perform software repository mining for bug analysis, or conduct empirical studies on software defects. Triggers when users ask to find bug-introducing commits, identify when a bug was introduced, trace bug origins, perform SZZ analysis, or analyze bug-fixing commits.
Instruments code to track the flow of untrusted or sensitive data at runtime, enabling detection of injection vulnerabilities, data leaks, and privilege violations. Use when users need to: (1) Track untrusted input propagation through code, (2) Detect SQL injection, XSS, or command injection vulnerabilities, (3) Identify sensitive data leaks, (4) Monitor privilege escalation paths, (5) Perform dynamic taint analysis for security testing. Supports Python, Java, JavaScript, and C/C++ with configurable taint sources and sinks.
Generate boilerplate code and project templates/skeletons automatically. Use when: (1) Creating new projects from scratch (React app, FastAPI backend, Express API), (2) Generating repetitive code patterns (CRUD endpoints, models, controllers), (3) Scaffolding components, services, or modules, (4) Creating test boilerplate, (5) Setting up monorepo structures. Provides project templates and code generation patterns for common development tasks.
Automatically updates a codebase to a new language version, framework version, or library update while ensuring all tests still pass. Use this skill when migrating Python 2→3, upgrading framework versions (React, Django, Angular), updating dependencies, or performing any version migration where tests define correct behavior. The skill analyzes failing tests caused by migration, categorizes errors (import errors, API changes, type errors, behavior changes), proposes systematic fixes, and verifies that test-observable behavior remains unchanged. Triggers when users ask to migrate code versions, upgrade dependencies while keeping tests passing, update framework versions, or perform test-guided version migrations.
Removes unnecessary code from a repository while preserving exactly the behavior exercised by a given test suite. Use this skill when you need to debloat, slim down, or remove unused code from a codebase based on test coverage. The skill analyzes which code elements (files, classes, functions, methods, branches, imports) are exercised by tests, identifies unreachable or unused code, and safely removes it while ensuring all tests continue to pass. Triggers when users ask to remove unused code, debloat based on tests, eliminate dead code guided by test coverage, or slim down a repository to only test-required functionality.
Scan repositories for newly disclosed CVEs in dependencies after a specific cutoff date. Takes a repository path, cutoff date (YYYY-MM-DD), and optional parameters for transitive dependencies. Parses dependency manifests (package.json, pom.xml, requirements.txt, go.mod, Cargo.toml) and lockfiles to extract exact versions. Queries vulnerability databases (OSV.dev, NVD, GitHub Advisory) to identify CVEs disclosed strictly after the cutoff date. Distinguishes between newly disclosed CVEs and previously known CVEs. Use when: (1) Performing security audits to find new vulnerabilities since last review, (2) Checking if new CVEs affect a historical codebase version, (3) Generating compliance reports showing vulnerability status at specific dates, (4) Tracking security posture changes over time. Supports npm, Maven, pip, Go modules, Cargo, and other major ecosystems.
Automatically repair C/C++ code violations detected by TLA+ model checking. Takes a program, TLA+ specification, and TLC counterexample trace as input, then generates minimal code modifications to eliminate the violation. Use when: (1) TLC model checker reports an invariant violation, deadlock, or temporal property failure, (2) You have a counterexample trace and need to fix the corresponding code, (3) You need to understand how a TLA+ violation maps to program-level bugs, (4) You want to validate repairs by re-running TLC. Supports safety properties (invariants), liveness properties (temporal logic), and deadlock detection.
Collect, normalize, and structure execution traces from instrumented programs (strace, ltrace) into JSON format for downstream analysis. Use when working with system call traces, library call traces, or execution logs that need to be analyzed for debugging, test case reproduction, or verification. Supports parsing strace/ltrace output, filtering noise, extracting debug information, and preparing traces for bug analysis or reproduction workflows.
Automatically generate TLA+ specifications from source code (C/C++, Python) for formal verification of distributed systems. Use when users need to: (1) Generate TLA+ specs from program implementations, (2) Model distributed systems, consensus protocols, or concurrent algorithms, (3) Extract state variables, actions, and invariants from code, (4) Create formal specifications for model checking with TLC, (5) Verify safety and liveness properties of distributed systems. Particularly effective for message-passing systems, replication protocols, consensus algorithms, and distributed transactions.
Automatically generates comprehensive unit tests for functions, classes, and modules. Use when you need to create tests for Python (pytest, unittest) or Java (JUnit, TestNG) code. Generates tests with comprehensive coverage including happy paths, edge cases, and error conditions. Analyzes existing test patterns in the codebase to match style and conventions. Supports mocking, parameterized tests, fixtures, and follows best practices for each framework.
Extract language-agnostic pseudocode from formally verified programs (Isabelle/HOL, Coq) while preserving verified control flow, data dependencies, and algorithmic logic. Use when: (1) Users have verified code and need readable pseudocode, (2) Documenting verified algorithms for broader audiences, (3) Translating verified implementations to other languages, (4) Creating algorithm specifications from verified code, (5) Preserving verification guarantees in pseudocode form, or (6) Abstracting proof-heavy code to essential logic. Maintains semantic faithfulness to verified implementation.
Establish explicit traceability between formal specifications (preconditions, postconditions, invariants) and verified code components with their correctness proofs. Produce structured Markdown mapping reports showing verification coverage and proof evidence. Use when auditing formal verification, documenting verified systems, establishing traceability for certification, or when the user asks to map specifications to code, generate verification reports, or analyze verification coverage in Coq, Dafny, Isabelle, or other proof assistants.
|- Profile Dart command-line applications using the VM Service protocol to capture CPU samples and identify performance bottlenecks. Helps agents automate CPU profiling, generate function call breakdown summaries, and export JSON profiles without a browser or DevTools.
Preserves or recovers the reasoning behind a codebase - architectural decisions, rejected alternatives, workarounds, incident learnings, operational constraints, and historical context the code itself cannot explain. Use when implementing or reviewing a non-trivial change involving a design decision, workaround, incident fix, operational constraint, rejected alternative, or changed assumption; when documenting an existing or legacy codebase; during onboarding or a maintainer handover; or when interviewing a developer before their knowledge is lost (e.g. before they leave or retire); or when the user expresses frustration with, or reports a problem with, this skill itself. Identifies what the code cannot explain, asks focused questions instead of generic ones, and maintains concise, topic-based, version-controlled documentation readable by both humans and AI agents.
Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.
The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase through parallel scouts, drafts every surviving project-state doc, resolves gaps with the user, and locks the project as initialized.
Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report, mentions ADR conflicts, or asks which downstream artifacts the current state supports. Never decides alone — every arbitration is user-attributed and logged.
The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk so a context reset loses nothing, then populates .codearbiter/ and locks it initialized. No project-state doc is written before the layers are solid; orchestration does not resume until the lock is set.
The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" and surfaces the merge decision to the user.
Finish an already-merged branch. Proves the branch is an ancestor of the fetched default, classifies leftover artifacts as unique / redundant / superseded, returns to a clean --ff-only default checkout, and deletes the merged local branch — every discard confirmed per item. Routed to by {{CMD:cleanup}}.
The secret-source gate. Routed to when changed code reads, writes, or passes a secret — API key, token, password, connection string, signing key, certificate, or any value that grants access. Validates that every secret comes from the approved store and never lands in source, log, test fixture, error response, image, or LLM prompt. The auth-crypto-reviewer agent is dispatched as the reviewer.
The single permitted path to a version tag, for any target declared in the project's .codearbiter/release-targets.md. Routed to when the user invokes /release on a non-default branch with a green suite. Takes the declared target as its one argument, derives the SemVer bump from Conventional-Commits history since that target's last tag, rolls the commits into that target's CHANGELOG, writes an annotated tag in that target's namespace, and on authorization publishes it as a GitHub Release with the changelog section as its notes. A release commit, if needed, routes through commit-gate; the tag and Release are never published without explicit authorization.
OPTIONAL per-task isolation for autonomous parallel work. Routed to only on explicit opt-in by subagent-driven-development or dispatching-parallel-agents, so parallel units mutate files without colliding. Stands up one worktree per unit, works it in isolation, then integrates each unit back onto the caller's working branch for the caller's single commit-gate + finishing-a-development-branch exit. Never on the default path; it does not bypass a gate or finish per unit.
The deep, rarely-convened whole-codebase audit lane. Routed to when the user invokes {{CMD:tribunal}}. Seven gated phases — cost/model, map, roster dispatch, triage, report, approval+filing, telemetry. Costs on the order of millions of tokens; proceeds only after the user acknowledges the estimate; never a required gate; nothing filed or sent without explicit authorization.
The brownfield back-fill. Routed to by /create-context, and by startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker but source code exists. Six gated phases — pre-flight, scout dispatch, synthesis, gap interview, write, lock. Reads the existing codebase through parallel scouts, drafts every surviving project-state doc, resolves gaps with the user, and locks the project as initialized.
Investigate-then-decide root-cause analysis for a defect whose cause is unknown (distinct from /fix, which assumes a known bug). Five gated phases: capture, hypothesize, gather, decide, hand off. Investigation only, no code changes; exits to /fix, /adr, or a no-action close.
Reconcile the project's architectural artifacts against the scaffold and prior decisions, then present each variance as a SMARTS analysis for the user to decide. Routed to when the user asks to arbitrate, reconcile, or consolidate architectural context, requests a variance report, mentions ADR conflicts, or asks which downstream artifacts the current state supports. Never decides alone — every arbitration is user-attributed and logged.
The greenfield decomposition interview. Routed to at startup when .codearbiter/CONTEXT.md lacks the <!--INITIALIZED--> body marker and no source code exists, or when the user invokes /decompose. A senior-architect persona drives a six-layer interview, persists every layer to disk so a context reset loses nothing, then populates .codearbiter/ and locks it initialized. No project-state doc is written before the layers are solid; orchestration does not resume until the lock is set.
The terminal step of /feature and /sprint. Routed to once commit-gate has cleared, to decide the branch's fate — merge via PR, open a PR, or discard. Direct merge to the default branch is forbidden; every change lands through a PR. Under /sprint the skill auto-selects "open PR" and surfaces the merge decision to the user.