Capabilities Reference
Quick lookup table of all skills, commands, agents, and hooks in the Claude Night Market.
For full flag documentation and workflow examples: See Capabilities Reference Details.
Quick Reference Index
All Skills (Alphabetical)
| Skill | Plugin | Description |
|---|---|---|
agent-expenditure | conserve | Per-agent token usage tracking |
additive-bias-defense | leyline | Scrutiny questions to challenge unnecessary additions |
agent-teams | conjure | Coordinate Claude Code Agent Teams through filesystem-based protocol |
api-review | pensive | API surface evaluation |
architecture-aware-init | attune | Architecture-aware project initialization with research |
architecture-diagram | cartograph | Component relationship diagrams |
architecture-paradigm-client-server | archetypes | Client-server communication |
architecture-paradigm-cqrs-es | archetypes | CQRS and Event Sourcing |
architecture-paradigm-domain-driven | archetypes | Modeling a business in its own language |
architecture-paradigm-event-driven | archetypes | Asynchronous communication |
architecture-paradigm-functional-core | archetypes | Functional Core, Imperative Shell |
architecture-paradigm-hexagonal | archetypes | Ports & Adapters architecture |
architecture-paradigm-layered | archetypes | Traditional N-tier architecture |
architecture-paradigm-microkernel | archetypes | Plugin-based extensibility |
architecture-paradigm-microservices | archetypes | Independent distributed services |
architecture-paradigm-modular-monolith | archetypes | Single deployment with internal boundaries |
architecture-paradigm-pipeline | archetypes | Pipes-and-filters model |
architecture-paradigm-serverless | archetypes | Function-as-a-Service |
architecture-paradigm-service-based | archetypes | Coarse-grained SOA |
architecture-paradigm-space-based | archetypes | Data-grid architecture |
architecture-paradigms | archetypes | Orchestrator for paradigm selection |
architecture-review | pensive | Architecture assessment |
authentication-patterns | leyline | Auth flow patterns |
blast-radius | pensive | Code change blast radius analysis with risk scoring |
bloat-detector | conserve | Detection algorithms for dead code, God classes, documentation duplication |
browser-recording | scry | Playwright browser recordings |
assisted-mastery | imbue | Assistance-dilemma resolution: visible reasoning, tradeoff surfacing, and fading help to build judgment |
graduated-implementation | imbue | Bounded start then ramp the next increment’s ambition only on demonstrated competence of the prior one |
bug-review | pensive | Bug hunting |
call-chain | cartograph | Trace execution paths through code knowledge graph |
catchup | imbue | Context recovery |
challenge | gauntlet | Adaptive difficulty challenge session for codebase knowledge testing |
class-diagram | cartograph | Class and interface diagrams |
clear-context | conserve | Auto-clear workflow with session state persistence |
code-communities | cartograph | Detect architectural clusters via community detection |
code-quality-principles | conserve | Core principles for AI-assisted code quality |
code-refinement | pensive | Duplication, algorithms, and clean code analysis |
code-search | tome | GitHub implementation search |
commit-messages | sanctum | Conventional commits |
compression-strategy | conserve | Context compression analysis and recommendations |
computer-control | phantom | Desktop automation via Claude’s vision and action API |
content-sanitization | leyline | External content sanitization |
context-map | conserve | Pre-scan project structure to reduce exploration token waste |
context-optimization | conserve | MECW principles and 50% context rule |
cpu-gpu-performance | conserve | Resource monitoring and selective testing |
curate | gauntlet | Add or edit knowledge annotations with tribal context |
damage-control | leyline | Agent crash recovery and state reconciliation |
data-flow | cartograph | Data movement diagrams |
decision-journal | leyline | Contract for the project decision journal (tradeoffs and lessons-learned logs) |
decisive-action | conserve | Decisive action patterns for efficient workflows |
deferred-capture | leyline | Contract for unified deferred-item capture across plugins |
delegation-core | conjure | Framework for delegation decisions |
dependency-graph | cartograph | Import and dependency diagrams |
dependency-verification | imbue | Package-hallucination and slopsquat defense: verifies a package exists before install |
diff-analysis | imbue | Semantic changeset analysis |
dig | tome | Interactive research refinement |
digital-garden-cultivator | memory-palace | Digital garden maintenance |
discourse | tome | Community discussion scanning |
do-issue | sanctum | GitHub issue resolution workflow |
doc-consolidation | sanctum | Document merging |
doc-generator | scribe | Generate and remediate documentation |
doc-importer | scribe | Import external documents to markdown |
doc-updates | sanctum | Documentation maintenance |
document-conversion | leyline | Universal document-to-markdown conversion |
dora-metrics | minister | Compute DORA delivery-performance metrics with tier classification |
dorodango | attune | Iterative code polishing workflow |
elegant-code | conserve | Minimal-code decision ladder with full safety, edge, and negative-case coverage |
error-patterns | leyline | Standardized error handling |
escalation-governance | abstract | Model escalation decisions |
evaluation-framework | leyline | Decision thresholds |
extract | gauntlet | Analyze codebase and build a knowledge base |
feature-review | imbue | Feature prioritization with RICE/WSJF/Kano scoring and optional research enrichment via tome (--research) |
file-analysis | sanctum | File structure analysis |
friction-detector | abstract | Detect friction signals and graduate recurring patterns into rules |
gauntlet-curate | gauntlet | Research and refresh the problem bank; surveys coverage gaps and proposes YAML-valid entries |
gemini-delegation | conjure | Gemini CLI integration |
gif-generation | scry | GIF processing and optimization |
git-platform | leyline | Cross-platform git forge detection and command mapping |
git-workspace-review | sanctum | Repo state analysis |
github-initiative-pulse | minister | Initiative progress tracking |
graph-build | gauntlet | Build or update the code knowledge graph |
graph-search | gauntlet | FTS5 search of the code knowledge graph |
harden | pensive | Active codebase hardening with NIST/CWE-cited findings and concrete remediation proposals |
hook-authoring | abstract | Security-first hook development |
hook-scope-guide | abstract | Decide where to place hooks (plugin/project/global) |
hooks-eval | abstract | Hook security scanning |
ideate | tome | Diverse ideation methods with category-spanning selection and rotation |
install-watchdog | egregore | Install crash-recovery watchdog |
justify | imbue | Anti-additive-bias change audit |
karpathy-principles | imbue | Compact four-principle synthesis for LLM coding pitfalls |
knowledge-intake | memory-palace | Intake and curation |
knowledge-locator | memory-palace | Spatial search |
memory-clarity-probe | memory-palace | Memory clarity assessment via anchor questions |
latent-space-engineering | imbue | Agent behavior shaping through instruction framing |
loop-optimization | leyline | Hand-vs-compiler decision rule for loop transforms |
makefile-generation | attune | Generate language-specific Makefiles |
makefile-review | pensive | Makefile best practices |
markdown-formatting | leyline | Line wrapping and style conventions |
math-review | pensive | Mathematical correctness |
mcp-code-execution | conserve | MCP patterns for data pipelines |
performance-review | pensive | Time and space complexity hotspot detection |
media-composition | scry | Multi-source media stitching |
memory-palace-architect | memory-palace | Building virtual palaces |
metacognitive-self-mod | abstract | Hyperagents self-improvement analysis |
methodology-curator | abstract | Surface expert frameworks for skill development |
mission-orchestrator | attune | Unified lifecycle orchestrator for project development |
modular-skills | abstract | Modular design patterns |
onboard | gauntlet | Guided five-stage onboarding path through a codebase |
papers | tome | Academic literature search |
plugin-review | abstract | Tiered plugin quality review with dependency-aware scoping |
pr-prep | sanctum | PR preparation |
pr-review | sanctum | PR review workflows |
precommit-setup | attune | Set up pre-commit hooks |
progressive-loading | leyline | Dynamic content loading |
project-brainstorming | attune | Socratic ideation workflow |
project-execution | attune | Systematic implementation |
project-init | attune | Interactive project initialization |
project-planning | attune | Architecture and task breakdown |
project-specification | attune | Spec creation from brainstorm |
proof-of-work | imbue | Evidence-based work validation |
pytest-config | leyline | Pytest configuration patterns |
python-async | parseltongue | Async patterns |
python-packaging | parseltongue | Packaging with uv |
python-performance | parseltongue | Profiling and optimization |
python-testing | parseltongue | Pytest/TDD workflows |
quality-gate | egregore | Pre-merge quality validation for autonomous sessions |
quota-management | leyline | Rate limiting and quotas |
qwen-delegation | conjure | Qwen MCP integration |
release-health-gates | minister | Release readiness checks |
research | tome | Multi-source research orchestration |
response-compression | conserve | Response compression patterns |
palace-diagram | memory-palace | Visual palace structure diagrams |
palace-index-curator | memory-palace | Web-capture index curation |
review-chamber | memory-palace | PR review knowledge capture and retrieval |
review-core | imbue | Scaffolding for detailed reviews |
rigorous-reasoning | imbue | Anti-sycophancy guardrails |
risk-classification | leyline | Inline 4-tier risk classification for agent tasks |
rule-catalog | hookify | Pre-built behavioral rule templates |
rules-eval | abstract | Evaluate and validate Claude Code rules in .claude/rules/ directories |
rust-review | pensive | Rust-specific checking |
safety-critical-patterns | pensive | NASA Power of 10 rules for robust code |
scope-guard | imbue | Anti-overengineering |
sem-integration | leyline | Semantic diff CLI detection and fallback |
service-registry | leyline | Service discovery patterns |
session-management | sanctum | Session naming, checkpointing, and resume strategies |
session-palace-builder | memory-palace | Session-specific palaces |
session-replay | scribe | Convert session JSONL into GIF/MP4/WebM replays via VHS |
session-to-post | scribe | Convert sessions into shareable blog posts or case studies |
setup | oracle | Install and configure the oracle ONNX inference daemon |
shared-patterns | abstract | Reusable plugin development patterns |
shell-review | pensive | Shell script auditing for safety and portability |
skill-authoring | abstract | TDD methodology for skill creation |
skill-graph-audit | abstract | Map Skill() refs across plugins; detect hubs, isolates, dangling targets |
skill-library-mission | attune | Build a project skill library via discovery, parallel authoring, and review |
skills-eval | abstract | Skill quality assessment |
slop-detector | scribe | Detect AI-generated content markers |
smart-sourcing | conserve | Balance accuracy with token efficiency |
spec-writing | spec-kit | Specification authoring |
speckit-orchestrator | spec-kit | Workflow coordination |
stack-create | sanctum | Initialize a branch stack from a multi-step plan |
stack-mode | sanctum | Shared stack detection and multi-PR iteration contract for /pr-review --stack and /fix-pr --stack |
stack-push | sanctum | Push stack branches and open or update dependent PRs |
stack-rebase | sanctum | Cascading rebase after a base PR merges |
stewardship | leyline | Cross-cutting stewardship principles with layer-specific guidance |
storage-templates | leyline | Storage abstraction patterns |
structured-output | imbue | Formatting patterns |
style-learner | scribe | Extract writing style from exemplar text |
subagent-testing | abstract | Testing patterns for subagent interactions |
summon | egregore | Spawn autonomous agent session with budget |
supply-chain-advisory | leyline | Known-bad version detection, lockfile auditing, incident response |
synthesize | tome | Research findings synthesis |
task-planning | spec-kit | Task generation |
tech-tutorial | scribe | Plan, draft, and refine technical tutorials |
test-review | pensive | Test quality review |
test-updates | sanctum | Test maintenance |
testing-quality-standards | leyline | Test quality guidelines |
tiered-audit | pensive | Three-tier escalation audit (git history, targeted, full) |
token-conservation | conserve | Token usage strategies |
triz | tome | TRIZ cross-domain analogical reasoning |
tutorial-updates | sanctum | Tutorial maintenance and updates |
unified-review | pensive | Review orchestration |
uninstall-watchdog | egregore | Remove crash-recovery watchdog |
update-readme | sanctum | README maintenance and updates |
usage-logging | leyline | Telemetry tracking |
utility | leyline | Utility-guided action selection for orchestration |
validate-pr | sanctum | Diff-derived PR test plan with revert-test quality checks |
version-updates | sanctum | Version bumping |
vhs-recording | scry | Terminal recordings with VHS |
voice-extract | scribe | SICO comparative extraction from writing samples |
voice-generate | scribe | Generate text in learned writing voice |
voice-learn | scribe | Learning loop from manual edits |
voice-review | scribe | Dual-gate review against voice profile |
vow-enforcement | imbue | Three-layer constraint enforcement with soft vows, hard vows, and external validators |
war-room | attune | Multi-LLM expert council with Type 1/2 reversibility routing |
war-room-checkpoint | attune | Inline reversibility assessment for embedded escalation |
workflow-diagram | cartograph | Process and state transition diagrams |
workflow-improvement | sanctum | Workflow retrospectives |
workflow-monitor | imbue | Workflow execution monitoring and issue creation |
workflow-setup | attune | Configure CI/CD pipelines |
writing-rules | hookify | Guide for authoring behavioral rules |
All Commands (Alphabetical)
| Command | Plugin | Description |
|---|---|---|
/acp | sanctum | Add, commit, push to current branch |
/aggregate-logs | abstract | Generate LEARNINGS.md from skill execution logs |
/ai-hygiene-audit | conserve | Audit codebase for AI-generated code quality issues (vibe coding, Tab bloat, slop) |
/analyze-skill | abstract | Skill complexity analysis |
/analyze-tests | parseltongue | Test suite health report |
/api-review | pensive | API surface review |
/architecture-review | pensive | Architecture assessment |
/attune:arch-init | attune | Initialize with architecture-aware templates |
/attune:blueprint | attune | Plan architecture and break down tasks |
/attune:brainstorm | attune | Brainstorm project ideas using Socratic questioning |
/attune:execute | attune | Execute implementation tasks systematically |
/attune:mission | attune | Run full project lifecycle as a single mission with state detection and recovery |
/attune:project-init | attune | Initialize project with development infrastructure |
/attune:specify | attune | Create detailed specifications from brainstorm |
/attune:upgrade-project | attune | Add or update configurations in existing project |
/attune:validate | attune | Validate project structure against best practices |
/attune:war-room | attune | Multi-LLM expert deliberation with reversibility-based routing |
/bloat-scan | conserve | Progressive bloat detection (3-tier scan) |
/bug-review | pensive | Bug hunting review |
/bulletproof-skill | abstract | Anti-rationalization workflow |
/catchup | imbue | Quick context recovery |
/check-async | parseltongue | Async pattern validation |
/close-issue | minister | Analyze if GitHub issues can be closed based on commits |
/commit-msg | sanctum | Generate commit message |
/context-report | abstract | Context optimization report |
/control-desktop | phantom | Run a computer use task on the desktop |
/create-command | abstract | Scaffold new command |
/create-hook | abstract | Scaffold new hook |
/create-issue | minister | Create GitHub issue with labels and references |
/create-skill | abstract | Scaffold new skill |
/create-tag | sanctum | Create git tags for releases |
/dismiss | egregore | Terminate autonomous agent session |
/do-issue | sanctum | Fix GitHub issues |
/doc-generate | scribe | Generate new documentation |
/doc-polish | scribe | Clean up AI-generated content |
/elegant-code-review | conserve | Review the working diff against the elegant-code decision ladder |
/evaluate-skill | abstract | Evaluate skill execution quality |
/filter-log | conserve | Suggest tier-1 filter commands for a log file before any compression or paste |
/fix-pr | sanctum | Address PR review comments |
/fix-workflow | sanctum | Workflow retrospective with automatic improvement context gathering |
/fixit | sanctum | Fix broken functionality from pasted output using research, TDD, and proof-of-work |
/full-review | pensive | Unified code review |
/garden | memory-palace | Manage digital gardens |
/gauntlet | gauntlet | Run an ad-hoc challenge session (5 questions, random scope) |
/gauntlet-curate | gauntlet | Add or edit a knowledge annotation |
/gauntlet-extract | gauntlet | Rebuild the knowledge base from the current codebase |
/gauntlet-graph | gauntlet | Build, search, and query the code knowledge graph |
/gauntlet-onboard | gauntlet | Start or resume a guided onboarding path |
/gauntlet-progress | gauntlet | Show challenge accuracy stats, weak areas, and streak |
/git-catchup | sanctum | Git repository catchup |
/harden | pensive | Active codebase hardening with NIST/CWE-cited findings and concrete remediation proposals |
/hookify | hookify | Create behavioral rules to prevent unwanted actions |
/hookify:configure | hookify | Interactive rule enable/disable interface |
/hookify:from-hook | hookify | Convert Python SDK hooks to declarative rules |
/hookify:help | hookify | Display hookify help and documentation |
/hookify:install | hookify | Install hookify rule from catalog |
/hookify:list | hookify | List all hookify rules with status |
/hooks-eval | abstract | Hook evaluation |
/improve-skills | abstract | Auto-improve skills from observability data |
/install-watchdog | egregore | Install crash-recovery watchdog |
/justify | imbue | Audit changes for additive bias |
/karpathy-check | imbue | Pre-flight gate for the four Karpathy principles |
/make-dogfood | abstract | Makefile enhancement |
/makefile-review | pensive | Makefile review |
/math-review | pensive | Mathematical review |
/merge-docs | sanctum | Consolidate ephemeral docs |
/performance-review | pensive | Time and space complexity hotspot review |
/navigate | memory-palace | Search palaces |
/optimize-context | conserve | Context optimization |
/oracle-setup | oracle | Install and configure the oracle ONNX inference daemon |
/palace | memory-palace | Manage palaces |
/plugin-review | abstract | Tiered plugin quality review (branch/pr/release) |
/pr-review | sanctum | Enhanced PR review |
/prepare-pr | sanctum | Complete PR preparation with updates and validation |
/promote-discussions | abstract | Promote highly-voted community learnings from Discussions to Issues |
/record-browser | scry | Record browser session |
/record-terminal | scry | Create terminal recording |
/refine-code | pensive | Analyze and improve living code quality |
/reinstall-all-plugins | leyline | Refresh all plugins |
/resolve-threads | sanctum | Resolve PR review threads |
/review-room | memory-palace | Manage PR review knowledge in palaces |
/rules-eval | abstract | Evaluate Claude Code rules for frontmatter, glob patterns, and content quality |
/run-profiler | parseltongue | Profile code execution |
/rust-review | pensive | Rust-specific review |
/session-replay | scribe | Generate GIF/MP4/WebM replay from session JSONL |
/session-to-post | scribe | Convert session into blog post or case study |
/shell-review | pensive | Shell script safety and portability review |
/skill-history | pensive | View recent skill executions with context |
/skill-library | attune | Build a project skill library as a resumable mission |
/skill-logs | memory-palace | View skill execution logs |
/skill-review | pensive | Analyze skill metrics and stability gaps |
/skills-eval | abstract | Skill quality assessment |
/speckit-analyze | spec-kit | Check artifact consistency |
/speckit-checklist | spec-kit | Generate checklist |
/speckit-clarify | spec-kit | Clarifying questions |
/speckit-constitution | spec-kit | Project constitution |
/speckit-converge | spec-kit | Append unbuilt work as tasks |
/speckit-implement | spec-kit | Execute tasks |
/speckit-plan | spec-kit | Generate plan |
/speckit-specify | spec-kit | Create specification |
/speckit-startup | spec-kit | Bootstrap workflow |
/speckit-tasks | spec-kit | Generate tasks |
/speckit-taskstoissues | spec-kit | Convert tasks.md entries to GitHub Issues |
/status | egregore | Check autonomous session status |
/stewardship-health | imbue | Display stewardship health dimensions for plugins |
/structured-review | imbue | Structured review workflow |
/style-learn | scribe | Create style profile from examples |
/summon | egregore | Spawn autonomous agent session with budget |
/sync-capabilities | sanctum | Detect and fix drift between plugin.json and docs |
/test-review | pensive | Test quality review |
/test-skill | abstract | Skill testing workflow |
/tome:cite | tome | Generate formatted bibliography |
/tome:dig | tome | Refine research results interactively |
/tome:export | tome | Export research findings |
/tome:research | tome | Run multi-source research session |
/unbloat | conserve | Safe bloat remediation with interactive approval |
/uninstall-watchdog | egregore | Remove crash-recovery watchdog |
/update-all-plugins | leyline | Update all plugins |
/update-ci | sanctum | Update pre-commit hooks and CI/CD workflows |
/update-dependencies | sanctum | Update project dependencies |
/update-docs | sanctum | Update documentation |
/update-labels | minister | Reorganize GitHub issue labels with professional taxonomy |
/update-plugins | sanctum | Audit plugin registrations + automatic performance analysis and improvement recommendations |
/update-tests | sanctum | Maintain tests |
/update-tutorial | sanctum | Update tutorial content |
/update-version | sanctum | Bump versions |
/validate-hook | abstract | Validate hook compliance |
/validate-plugin | abstract | Check plugin structure |
/validate-pr | sanctum | Diff-derived PR test plan with area-targeted checks and revert-test quality proof |
/verify-plugin | leyline | Verify plugin behavioral contract history via GitHub Attestations |
/visualize | cartograph | Generate codebase diagrams via Mermaid Chart MCP |
/voice-extract | scribe | Extract writing voice from samples |
/voice-generate | scribe | Generate text in trained voice |
/voice-learn | scribe | Learn from manual edits |
/voice-review | scribe | Review text against voice profile |
All Agents (Alphabetical)
| Agent | Plugin | Description |
|---|---|---|
ai-hygiene-auditor | conserve | Audit codebases for AI-generation warning signs |
architecture-reviewer | pensive | Principal-level architecture review |
blast-radius-reviewer | pensive | Graph-aware code review using blast radius analysis |
bloat-auditor | conserve | Orchestrates bloat detection scans |
code-refiner | pensive | Code quality refinement orchestrator |
code-reviewer | pensive | Expert code review |
code-searcher | tome | GitHub code search |
craft-reviewer | scribe | Writing craft evaluation (naming, structure, anchoring) |
codebase-explorer | cartograph | Codebase structure analysis for diagrams |
commit-agent | sanctum | Commit message generator |
context-optimizer | conserve | Context optimization |
continuation-agent | conserve | Continue work from session state checkpoint |
dependency-updater | sanctum | Dependency version management |
desktop-pilot | phantom | Autonomous desktop control via Computer Use API |
discourse-scanner | tome | Community discourse scanning |
doc-editor | scribe | Interactive documentation editing |
doc-verifier | scribe | QA validation using proof-of-work methodology |
extractor | gauntlet | Autonomous knowledge extraction agent for gauntlet knowledge base |
garden-curator | memory-palace | Digital garden maintenance |
git-workspace-agent | sanctum | Repository state analyzer |
harden-orchestrator | pensive | Active hardening orchestrator (composes rust-review, supply-chain-advisory, bug-review) |
implementation-executor | spec-kit | Task executor |
knowledge-librarian | memory-palace | Knowledge routing |
knowledge-navigator | memory-palace | Palace search |
literature-reviewer | tome | Academic literature review |
media-recorder | scry | Autonomous media generation for demos and GIFs |
meta-architect | abstract | Plugin ecosystem design |
orchestrator | egregore | Autonomous development lifecycle agent |
palace-architect | memory-palace | Palace design |
plugin-validator | abstract | Plugin validation |
pr-agent | sanctum | PR preparation |
prose-reviewer | scribe | AI patterns, banned phrases, voice drift detection |
research | tome | Multi-source research orchestrator (delegates to Skill(tome:research)) |
project-architect | attune | Guides full-cycle workflow (brainstorm to plan) |
project-implementer | attune | Executes implementation with TDD |
python-linter | parseltongue | Strict ruff linting without bypasses |
python-optimizer | parseltongue | Performance optimization |
python-pro | parseltongue | Python 3.9+ expertise |
python-tester | parseltongue | Testing expertise |
review-analyst | imbue | Structured reviews |
rust-auditor | pensive | Rust security audit |
sentinel | egregore | Watchdog agent for crash recovery |
skill-auditor | abstract | Skill quality audit |
skill-evaluator | abstract | Skill execution evaluator |
skill-improver | abstract | Implements skill improvements from observability |
insight-engine | abstract | Deep analysis for bugs, optimizations, and improvements |
slop-hunter | scribe | Full-document AI slop detection |
spec-analyzer | spec-kit | Spec consistency |
task-generator | spec-kit | Task creation |
triz-analyst | tome | TRIZ cross-domain analysis |
unbloat-remediator | conserve | Executes safe bloat remediation |
workflow-improvement-analysis-agent | sanctum | Workflow improvement analysis |
workflow-improvement-implementer-agent | sanctum | Workflow improvement implementation |
workflow-improvement-planner-agent | sanctum | Workflow improvement planning |
workflow-improvement-validator-agent | sanctum | Workflow improvement validation |
workflow-recreate-agent | sanctum | Workflow reconstruction |
All Hooks (Alphabetical)
| Hook | Plugin | Type | Description |
|---|---|---|---|
aggregate_learnings_daily.py | abstract | UserPromptSubmit | Daily learning aggregation (24h cadence) with severity-based issue creation |
auto-star-repo.sh | leyline | SessionStart | Auto-star the repo if not already starred |
config_change_audit.py | sanctum | ConfigChange | Audit configuration changes |
context_warning.py | conserve | PreToolUse | Context utilization monitoring |
daemon_lifecycle.py | oracle | SessionStart, Stop | Oracle daemon lifecycle management |
deferred_item_sweep.py | sanctum | Stop | Sweep session ledger and file deferred items as GitHub issues |
deferred_item_watcher.py | sanctum | PostToolUse | Detect deferred items in Skill output and write to session ledger |
detect-git-platform.sh | leyline | SessionStart | Detect git forge platform from remote URL |
fetch-recent-discussions.sh | leyline | SessionStart | Fetch recent GitHub Discussions |
graph_auto_update.py | gauntlet | PostToolUse | Auto-update code graph after git commits |
graph_community_refresh.py | cartograph | PostToolUse | Refresh community detection after graph builds |
homeostatic_monitor.py | abstract | PostToolUse | Stability gap monitoring, queues degrading skills for improvement |
local_doc_processor.py | memory-palace | PostToolUse | Processes local docs |
noqa_guard.py | leyline | PreToolUse | Block inline lint suppression directives |
permission_denied_logger.py | conserve | PermissionDenied | Log auto-mode permission denials for observability |
permission_request.py | conserve | PermissionRequest | Permission automation |
post-evaluation.json | abstract | Config | Quality scoring config |
post_implementation_policy.py | sanctum | SessionStart | Requires docs/tests updates |
post_learnings_stop.py | abstract | Stop | Post learnings to GitHub Discussions on session stop |
pr_blast_radius.py | pensive | PreToolUse | Surface blast radius context on PR creation |
pre-skill-load.json | abstract | Config | Pre-load validation |
pre_compact.py | tome | PreCompact | Checkpoint active research session |
pre_compact_preserve.py | conserve | PreCompact | Preserve critical context before compression |
pre_skill_execution.py | abstract | PreToolUse | Skill execution tracking |
precommit_gate.py | gauntlet | PreToolUse | Pre-commit quality gate for gauntlet |
research_interceptor.py | memory-palace | PreToolUse | Cache lookup before web |
sanitize_external_content.py | leyline | PostToolUse | Sanitize external content for prompt injection |
security_pattern_check.py | sanctum | PreToolUse | Security anti-pattern detection |
session-start.sh | conserve, imbue | SessionStart | Session initialization |
session_complete_notify.py | sanctum | Stop, UserPromptSubmit | Cross-platform toast notifications and state management |
session_lifecycle.py | memory-palace | Stop | Session lifecycle management |
session_start.py | tome | SessionStart | Check for active research sessions |
session_start_hook.py | egregore | SessionStart | Inject manifest context into new sessions |
setup.sh | conserve | Setup | Environment initialization |
setup.sh | memory-palace | Setup | Palace directory initialization |
skill_execution_logger.py | abstract | PostToolUse | Skill metrics logging |
stop_hook.py | egregore | Stop | Prevent early exit while work items remain |
supply_chain_check.py | leyline | SessionStart | Warn about known-compromised package versions in lockfiles |
task_created_tracker.py | sanctum | TaskCreated | Track task creation for workflow completeness monitoring |
tdd_bdd_gate.py | imbue | PreToolUse | Iron Law enforcement at write-time |
tool_output_summarizer.py | conserve | PostToolUse | Monitor and warn about tool output bloat |
url_detector.py | memory-palace | UserPromptSubmit | URL detection |
user-prompt-submit.sh | imbue | UserPromptSubmit | Scope validation |
vow_bounded_reads.py | imbue | PreToolUse | Warns when discovery read budget (15 reads) is exceeded per session |
vow_no_ai_attribution.py | imbue | PreToolUse | Blocks AI attribution strings (Co-authored-by: Claude, etc.) in commits |
vow_no_emoji_commits.py | imbue | PreToolUse | Blocks emoji characters in git commit messages |
user_prompt_hook.py | egregore | UserPromptSubmit | Resume orchestration after user interrupts |
verify_workflow_complete.py | sanctum | Stop | End-of-session workflow verification |
web_research_handler.py | memory-palace | PostToolUse | Web research processing and storage prompting; skips non-2xx error pages (#547) |