he-reconcile
Analyze repo, tracker, PR, validation, session, and .harness evidence. Use when multi-stage Harness Engineering work needs safe resume routing.
45
47%
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/skills/he-reconcile/SKILL.mdQuality
Discovery
59%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 targets a very specific niche (Harness Engineering resume routing) and includes both 'what' and 'when' clauses, which is good for completeness and distinctiveness. However, the actions described are vague ('analyze'), and the trigger terms are highly specialized jargon that users would rarely use naturally, limiting discoverability.
Suggestions
Replace vague 'analyze' with specific concrete actions (e.g., 'Parse .harness files, validate PR state, determine safe resume points for multi-stage pipelines').
Add natural-language trigger terms users might actually say, such as 'resume pipeline', 'continue harness run', 'pick up where I left off', or 'check pipeline state'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | It names a domain ('Harness Engineering') and lists types of evidence to analyze ('repo, tracker, PR, validation, session, .harness'), but the actual actions are vague—'analyze' and 'safe resume routing' don't describe concrete operations like 'parse', 'validate', 'merge', or 'generate reports'. | 2 / 3 |
Completeness | It does answer both 'what' (analyze repo, tracker, PR, validation, session, and .harness evidence) and 'when' (when multi-stage Harness Engineering work needs safe resume routing), with an explicit 'Use when...' clause. | 3 / 3 |
Trigger Term Quality | The terms used are highly specialized jargon ('harness evidence', 'safe resume routing', 'multi-stage Harness Engineering') that users would be unlikely to naturally say. Common variations or natural language triggers are absent. | 1 / 3 |
Distinctiveness Conflict Risk | The description is highly specific to a niche domain ('Harness Engineering', '.harness evidence', 'safe resume routing'), making it very unlikely to conflict with other skills. | 3 / 3 |
Total | 9 / 12 Passed |
Implementation
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads as an abstract coordination protocol rather than actionable guidance. It defines roles, boundaries, and routing rules clearly but fails to provide any concrete examples of what reconciliation actually looks like in practice—no sample output JSON, no real commands, no executable steps. The heavy reliance on unavailable reference files means much of the actual procedural content is deferred elsewhere.
Suggestions
Add a concrete, complete example of the output format (e.g., a sample JSON object with realistic field values for a common reconciliation scenario) to make the Output Format section actionable.
Replace the abstract procedure steps with specific, executable guidance—e.g., show the actual commands or file reads needed to 'reconstruct lifecycle state' from .harness evidence files.
Expand the Examples section with at least one full input-to-output walkthrough showing the stage map, blockers, and routing decision for a realistic scenario.
Consolidate the numerous reference file pointers into a single well-organized reference table rather than scattering 'See references/...' lines across multiple sections.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is moderately efficient but includes significant jargon-heavy prose that could be tightened. Phrases like 'Coordinate state, not ceremony' and repeated references to folded context files add overhead. Several sections (Inputs, Outputs, Output Format) list many fields without explaining their purpose, creating density without clarity. | 2 / 3 |
Actionability | The skill provides no executable code, no concrete commands, no specific examples of input/output structures, and no copy-paste ready guidance. The procedure is abstract ('Reconstruct lifecycle state from live repo, tracker, PR...') with no concrete steps showing how to actually do this. The output format lists field names but provides no schema, no example JSON, and no concrete values. | 1 / 3 |
Workflow Clarity | The procedure has a numbered sequence and the validation section mentions pass/fail/blocked gates, but critical steps are deferred to 'references/hot-path-folded-context.md' which is not provided. The feedback loop for failure handling exists conceptually (stop with blocked_reason and smallest recovery step) but lacks explicit validation checkpoints between procedure steps. | 2 / 3 |
Progressive Disclosure | The skill references multiple external files (contract.yaml, evals.yaml, hot-path-folded-context.md, several shared references) which shows intent for progressive disclosure, but none of the bundle files are provided, making it impossible to verify the references resolve. Multiple sections end with 'See references/hot-path-folded-context.md for folded X detail' which is good signaling but the repeated pattern and the sheer number of references (10+) creates navigation overhead. | 2 / 3 |
Total | 7 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- jscraik/Agent-Skills
- Commit
4c78f98
- 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.