he-work
Implement approved Harness Engineering work. Use when a plan, todo list, or tiny spec needs traceable delivery and validation.
46
48%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./Plugins/harness-engineering/fixtures/budget-archive/2026-04-21/deferred-store/skills/team_automation/he-work/SKILL.mdQuality
Discovery
40%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description is too vague about what the skill actually does—'implement approved Harness Engineering work' could mean almost anything. While it includes a 'Use when' clause and some trigger terms, the lack of concrete actions and the reliance on abstract phrases like 'traceable delivery and validation' make it difficult for Claude to confidently select this skill over others.
Suggestions
Replace 'implement approved Harness Engineering work' with specific concrete actions (e.g., 'Creates branches, writes code, runs tests, and submits PRs for approved Harness Engineering tasks').
Expand trigger terms with natural user phrases (e.g., 'Use when the user references a Harness ticket, approved plan, implementation checklist, or asks to execute a spec').
Clarify what 'traceable delivery and validation' means in practice—list the specific validation steps or deliverables the skill produces.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague language like 'implement approved work' and 'traceable delivery and validation' without listing any concrete actions. There are no specific capabilities described—what does 'implement' actually entail? No mention of specific file types, tools, or operations. | 1 / 3 |
Completeness | It has a 'Use when...' clause which addresses the 'when' question, but the 'what' is extremely vague ('implement approved Harness Engineering work' tells us almost nothing about concrete actions). The 'when' triggers are also somewhat abstract ('traceable delivery and validation'). | 2 / 3 |
Trigger Term Quality | Contains some relevant terms like 'plan', 'todo list', 'tiny spec' that a user might reference, but 'Harness Engineering' is domain-specific jargon that may not match natural user language. Missing common variations or more natural trigger phrases. | 2 / 3 |
Distinctiveness Conflict Risk | 'Harness Engineering' provides some domain specificity, but 'implement approved work' and 'plan, todo list' are generic enough to overlap with many implementation or task-management skills. The description doesn't carve out a clear, unique niche. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
57%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-organized process skill that effectively uses progressive disclosure to keep the overview lean while pointing to detailed references. Its main weaknesses are moderate redundancy across sections (Constraints, Gotchas, Anti-patterns overlap) and a lack of concrete, executable examples—most guidance is procedural description rather than specific commands or templates. The workflow is present but would benefit from integrated validation checkpoints rather than a separate validation section.
Suggestions
Consolidate overlapping content across Constraints, Gotchas, and Anti-patterns into a single 'Guardrails' section to eliminate redundancy and improve conciseness.
Add a concrete example of a verified slice workflow showing actual commands: e.g., a specific git, Linear CLI, or validation command sequence rather than abstract steps like 'Resolve Linear, branch, PR, IDs'.
Integrate validation checkpoints directly into the Procedure steps (e.g., after step 3, explicitly show running the lint script and the expected pass/fail decision) rather than keeping validation as a separate section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is reasonably efficient but has some redundancy—'Do not claim completion without validation evidence' appears in both Constraints and Gotchas, and several sections (Anti-patterns, Gotchas, Failure mode) overlap significantly. The 'Context disposition' line at the top is meta-instruction noise. However, it mostly avoids explaining things Claude already knows. | 2 / 3 |
Actionability | The skill provides one concrete executable command (the traceability lint script), but most guidance is procedural/conceptual rather than copy-paste ready. Steps like 'Resolve Linear, branch, PR, IDs, and validation gates' and 'Implement one verified slice at a time' are directional but lack specific commands or concrete examples of what a slice looks like or how to update Linear. | 2 / 3 |
Workflow Clarity | The 4-step Procedure section provides a clear sequence, and the Validation section includes a stop-on-failure gate. However, the procedure steps are high-level abstractions without explicit validation checkpoints between them—the validation section is separate rather than integrated into the workflow. The feedback loop (stop on drift/failed gates) is mentioned but not structured as a concrete validate-fix-retry cycle. | 2 / 3 |
Progressive Disclosure | The skill is well-structured as an overview with clearly signaled one-level-deep references to full guide, handoff/shipping, execution modes, approval flow, session evidence contract, subagent routing, and domain routing. The main body stays concise while pointing to detailed materials for deeper context. | 3 / 3 |
Total | 9 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_version | 'metadata.version' is missing | Warning |
Total | 10 / 11 Passed | |
- Repository
- jscraik/Agent-Skills
- Commit
d00c351
- Reviewed
Table of Contents
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.