he-heartbeat
Create bounded Harness Engineering follow-up checkpoints, resume packets, stop conditions, and automation-safe handoff notes. Use when work must continue later without losing state or silently becoming autonomous execution.
53
60%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./Plugins/harness-engineering/skills/he-heartbeat/SKILL.mdQuality
Discovery
85%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 well-structured with a clear what/when split and lists specific concrete deliverables. Its main weakness is that the trigger terms lean heavily on domain-specific jargon ('Harness Engineering', 'resume packets') that users may not naturally use when requesting this kind of help. The distinctiveness is excellent due to its narrow, well-defined niche.
Suggestions
Add more natural-language trigger terms users might say, such as 'pause work', 'save progress', 'pick up later', 'task handoff', or 'multi-session work'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'bounded follow-up checkpoints', 'resume packets', 'stop conditions', and 'automation-safe handoff notes'. These are distinct, named deliverables. | 3 / 3 |
Completeness | Clearly answers both what ('Create bounded follow-up checkpoints, resume packets, stop conditions, and automation-safe handoff notes') and when ('Use when work must continue later without losing state or silently becoming autonomous execution'). | 3 / 3 |
Trigger Term Quality | Includes some relevant terms like 'continue later', 'losing state', 'handoff', and 'autonomous execution', but the domain-specific jargon ('Harness Engineering', 'resume packets', 'stop conditions') may not match natural user language. Users might say things like 'pause work', 'pick up later', 'save progress', or 'hand off' which aren't well covered. | 2 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche focused on 'Harness Engineering' continuity artifacts with specific concepts like 'stop conditions' and 'automation-safe handoff notes'. Unlikely to conflict with other skills due to its specialized domain. | 3 / 3 |
Total | 11 / 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 provides a reasonable structural framework for heartbeat/checkpoint management but suffers significantly from lack of actionability—there are no concrete templates, schemas, or executable examples that would let Claude produce correct output. The workflow is sequenced but vague, and there is notable redundancy between Gotchas, Anti-Patterns, and Constraints sections. The skill would benefit greatly from a concrete heartbeat prompt template and output schema example.
Suggestions
Add a concrete heartbeat output template/schema showing the exact JSON structure with fields like schema_version, next_invocation, stop_rule, subagent_policy, etc., so Claude can produce correct structured output.
Replace the natural language examples with full input→output examples showing a user request and the resulting heartbeat configuration/prompt that Claude should generate.
Consolidate the overlapping Gotchas, Anti-Patterns, and Constraints sections into a single 'Guardrails' section to reduce redundancy and improve token efficiency.
Add a concrete validation checkpoint example (e.g., a checklist or specific checks to run) rather than the abstract 'Confirm schedule, destination, and safe prompt scope' instruction.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is moderately efficient but includes some sections that feel redundant or could be tightened. For example, 'Gotchas' and 'Anti-Patterns' overlap significantly (duplicate automations, stop rules), and some phrasing like 'Fail fast: stop at the first failed gate and do not proceed' is somewhat verbose for what it conveys. The philosophy section is brief but adds little actionable value. | 2 / 3 |
Actionability | The skill provides mostly abstract, procedural descriptions rather than concrete, executable guidance. There are no code snippets, command examples, JSON schemas, or copy-paste-ready templates. The 'Outputs' section mentions a schema_version and fields but never shows the actual schema. The 'Examples' section provides natural language descriptions of use cases but no actual heartbeat prompt templates or structured output examples. | 1 / 3 |
Workflow Clarity | The Procedure section provides a numbered sequence of steps, and there are validation and failure mode sections that mention stopping at failed gates. However, the steps are vague ('Encode cadence, destination, stop criteria'), lack concrete validation checkpoints with specific commands or checks, and the feedback loop for error recovery is underspecified ('stop and return the blocker with the smallest recovery step'). | 2 / 3 |
Progressive Disclosure | The References section provides clear one-level-deep links to supporting documents, which is good structure. However, no bundle files were provided to verify these references exist, and the main content itself could benefit from better separation—the Execution Boundaries and Constraints sections contain dense policy information that might be better as referenced documents. The overall organization is reasonable but the inline content is somewhat monolithic. | 2 / 3 |
Total | 7 / 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_field | 'metadata' should map string keys to string values | 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.