tk-logl

tk-logl/sentinel

Plugin Claude Code

Universal AI code quality enforcement — prevents 47 common AI mistakes.

tk-logl/sentinel

Plugin Claude Code

Universal AI code quality enforcement — prevents 47 common AI mistakes, enforces surgical edits, blocks dummy code, manages context memory, task lifecycle tracking, context-aware analysis.

tk-logl/sentinel

Agent

Use this agent for code quality review that checks against 47 AI mistake patterns. Use when reviewing code changes, PRs, or completed implementations. Detects false completions, abandoned test code, scope reduction, and all common AI coding anti-patterns.

tk-logl/sentinel

Agent

Use this agent to verify task completion with evidence. Checks that tests actually pass, builds succeed, no dummy code remains, no unchecked items, and the full scope was delivered without reduction. Prevents false completion claims.

tk-logl/sentinel

Command

Run a comprehensive rule compliance check on the current project.

tk-logl/sentinel

Command

Generate or update a file header comment for the specified file.

tk-logl/sentinel

Command

Initialize sentinel in the current project — creates .sentinel/ directory and configuration.

tk-logl/sentinel

Command

Run targeted mutation testing on specified files to validate test quality.

tk-logl/sentinel

Hook

Runs when you submit a prompt, before the agent sees it, executing scope-guard.sh and task-scope-guard.sh via bash (2 commands). From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs when a session starts, executing session-init.sh via bash. From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs before the agent uses a tool for Write, Edit, Bash and Task tool calls, executing pre-edit-gate.sh, deny-dummy.sh, surgical-change.sh, secret-scan.sh, scope-reduction-guard.sh, tdd-enforce.sh, pre-write-ast.sh, env-safety.sh and subagent-context.sh via bash (9 commands). From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs after a tool call finishes for Write, Edit, Bash and TaskUpdate tool calls, executing post-edit-verify.sh, file-header-check.sh, ast-quality-gate.py, ast-quality-gate-ts.js, post-edit-lint.sh, ast-grep-gate.sh, spec-auto-test.sh, error-logger.sh, task-automark.sh and task-completion-gate.sh via bash (10…

tk-logl/sentinel

Hook

Runs when a tool call fails, executing error-logger.sh via bash. From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs before the context is compacted, executing state-preserve.sh via bash. From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs when the agent finishes a response, executing completion-check.sh and spec-verify.sh via bash (2 commands). From tk-logl/sentinel.

tk-logl/sentinel

Hook

Runs when a session ends, executing session-save.sh via bash. From tk-logl/sentinel.

tk-logl/sentinel

Skill Claude CodeCodex

Reference guide for 47 common AI coding mistake patterns. Use when you need to check code against known anti-patterns, or when reviewing changes for quality issues.

tk-logl/sentinel

Skill Claude CodeCodex

Pre-implementation checklist workflow. Use before writing any source code to analyze impact, choose approach, and document blast radius. This is your blueprint — fill it with real analysis.

tk-logl/sentinel

Skill Claude CodeCodex

Generate and manage file header comments for key source files. Headers describe the file's role, co-modify files, invariants, and verification commands. Use for files over 200 lines.

tk-logl/sentinel

Skill Claude CodeCodex

Protect against context loss during compaction and session changes. Structured state preservation with 5-section format and multi-location recovery chain.