agent-hooks

Four lifecycle hooks every agent runs: on-session-start scans LESSONS-LEARNED.md, resumes checkpoints; on-pre-delegate verifies tracker issues, file partitions; on-post-delegate runs fast-review, CI checks; on-session-end runs the health checks, writes logs. Use when starting a new session, running pre-flight checks before delegation, coordinating between agents, reviewing a completed handoff, or wrapping up a session. Trigger terms: multi-agent setup, delegate tasks, agent coordination, session management, run pre-flight checks, start a new session, coordinate between agents, wrap up session

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A compact, action-oriented hook reference with strong sequencing and explicit gating for the delegation flow. Its main weakness is progressive disclosure: it repeatedly points to HOOKS-REFERENCE.md for the details that justify the lean main file, but that file is absent.

Suggestions

Add the missing HOOKS-REFERENCE.md bundle file containing the extended startup checks, approval polling, skill-matrix verification, and per-check example commands the body advertises, or inline that detail and drop the dangling links.

Make the deferred verification commands (fast-review invocation, the codebase-tool lint/typecheck/test commands, AC-verification steps) concrete in the body rather than pointing to the absent reference.

Add an explicit re-validate/re-delegate feedback loop note under on-post-delegate (the 3rd-failure path is named but the retry→re-review cycle is only implied).

Dimension	Reasoning	Score
Conciseness	The body is lean and table-driven with no padding or explanation of concepts Claude already knows; every line earns its place (lifecycle diagram, numbered action tables, two anti-patterns).	5 / 5
Actionability	It gives concrete, copy-pasteable commands ('rg -n "keyword" .opencastle/LESSONS-LEARNED.md', 'gh issue view TAS-XX', 'comm -12 <(sort agent1-files) <(sort agent2-files)', 'opencastle log --type=...'), but a few entries are pointers ('Run fast-review skill', 'Lint, typecheck, and test ... via the codebase-tool slot') rather than executable steps, leaving minor gaps.	4 / 5
Workflow Clarity	Each hook is a numbered sequence with a lifecycle diagram, and on-pre-delegate has an explicit hard gate ('All 6 must pass before the sub-agent is dispatched') plus on-session-end's HARD GATE; validation is largely present (CI checks, AC verification, '3rd failure → AGENT-FAILURES.md'), though some checkpoints are implicit and detailed commands are deferred to a missing reference file.	4 / 5
Progressive Disclosure	Sections are well organized and a one-level-deep reference (HOOKS-REFERENCE.md) is clearly signaled four times, but that referenced file does not exist in the bundle — so the advertised navigation leads nowhere and the teased extended content (approval polling, per-check example commands) is unreachable, undercutting the disclosure structure.	3 / 5
	Total	16 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that names concrete per-hook actions and provides explicit 'Use when' guidance plus a dedicated trigger-terms list. The only soft spot is trigger-term phrasing that is good but not maximally natural across synonyms.

Suggestions

Trim redundancy between the 'Use when' list and the 'Trigger terms' list so the natural phrasings users would actually say are not diluted by near-duplicates.

Add the most common lay phrasings (e.g. 'hand off work to another agent', 'check work before delegating') alongside the existing jargon-leaning trigger terms.

Dimension	Reasoning	Score
Specificity	The description enumerates concrete actions per hook — 'scans LESSONS-LEARNED.md', 'resumes checkpoints', 'verifies tracker issues, file partitions', 'runs fast-review, CI checks', 'runs the health checks, writes logs' — giving comprehensive, non-abstract coverage of what each lifecycle hook does.	5 / 5
Completeness	It explicitly answers both 'what' (four named hooks with concrete actions) and 'when' ('Use when starting a new session, running pre-flight checks before delegation...'), with concrete trigger phrases — matching the top anchor.	5 / 5
Trigger Term Quality	An explicit 'Trigger terms:' list supplies natural phrases ('multi-agent setup', 'delegate tasks', 'session management', 'start a new session', 'wrap up session'), but several overlap the 'Use when' list and a few lean jargon-flavored rather than the most common user phrasing, leaving minor gaps.	4 / 5
Distinctiveness Conflict Risk	It carves a clear niche — agent lifecycle hooks in a multi-agent opencastle workflow — with distinct, narrow triggers and minimal overlap risk against unrelated skills.	5 / 5
	Total	19 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 3 missing	Warning

	Total	15 / 16 Passed

Repository: monkilabs/opencastle
Path: src/orchestrator/skills/agent-hooks/SKILL.md
Commit: f297837

Reviewed: 6 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.