CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

hermes-ephemeral-delegation

Trigger: a mapping need, parallel units, context backstop, high-risk verify, fresh review, or multi-step debug. Delegate via delegate_task for a named reason.

Quality

75%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./internal/assets/skills/hermes-ephemeral-delegation/SKILL.md
SKILL.md
Quality
Evals
Security

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.

The body is exceptionally concise with a well-sequenced workflow and a clean, verified one-level reference structure. Its main weakness is actionability: several mandatory mission inputs cite sections and gates (Implementation Routing, Delegated Verification Gate) that are not defined anywhere in the skill or its bundle.

Suggestions

Define or link the missing prerequisites: either add the `## Implementation Routing` section and the Delegated Verification Gate handoff to this skill (or its references), or point to the exact file/skill where they live.

Add an explicit error-recovery step after verification, e.g. "If verification fails, re-delegate with the discrepancy quoted in the mission or fix inline, then re-verify".

Briefly define load-bearing local terms like "inline evidence budget" and "logbook" in one line each so the gates are executable without outside context.

DimensionReasoningScore

Conciseness

The body is lean and dense with zero padding: no basic concepts are explained, and sections like Hard Rules and the Decision Gates table convey rules with minimal tokens, matching "every token earns its place".

5 / 5

Actionability

The mission template and decision table are concrete, but step 2 mandates "the applicable test-first policy and runner from `## Implementation Routing`" and "the Verify handoff from the Delegated Verification Gate" — sections that do not exist in this skill or its bundle, leaving key details unresolvable. "Inline evidence budget" and "your logbook" are similarly undefined.

3 / 5

Workflow Clarity

The six Execution Steps are clearly sequenced with an explicit verification checkpoint (step 5) reinforced by the self-report verification rule, but there is no explicit error-recovery loop for when verification fails, keeping it below the score-5 anchor.

4 / 5

Progressive Disclosure

A single one-level-deep reference (references/tuning-knobs.md, verified to exist) is clearly signaled with a description of its contents, and the config-knob table is correctly split out of the main body; the body's sections are well-organized and easy to navigate.

5 / 5

Total

17

/

20

Passed

Description

58%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.

The description has an explicit trigger structure and names its tool, but its vocabulary is internal shorthand rather than natural user phrasing, weakening trigger quality and specificity. It is distinguishable as a delegation skill yet overlaps with dedicated review/debug skills on some triggers.

Suggestions

Rewrite trigger terms as natural user phrases, e.g. "Use when the user asks to parallelize independent work, delegate a code review to a fresh reviewer, or keep debugging output out of the main context".

Expand the 'what' beyond one action: state that the skill spawns ephemeral worker subagents, passes them self-contained missions, and verifies their self-reported results.

Add plain-language equivalents for jargon terms like "context backstop", "mapping need", and "parallel units" so the description reads clearly outside its own suite.

DimensionReasoningScore

Specificity

The only concrete action stated is "Delegate via delegate_task for a named reason"; the rest ("mapping need, parallel units, context backstop, high-risk verify") are compressed situation labels rather than specific capabilities, matching the 'names domain and 1-2 concrete actions' anchor.

3 / 5

Completeness

Both what ("Delegate via delegate_task") and an explicit when ("Trigger: a mapping need, parallel units...") are present, but the trigger clause is cryptic shorthand that could be more explicit and specific, fitting the score-4 anchor.

4 / 5

Trigger Term Quality

Some relevant keywords exist (delegate, parallel, review, debug), but coinages like "context backstop", "parallel units", and "high-risk verify" are internal jargon no user would naturally say, and common variations are missing.

3 / 5

Distinctiveness Conflict Risk

Naming the delegate_task tool carves out a delegation niche, but "fresh review" and "multi-step debug" would naturally fire for dedicated review and debugging skills, leaving overlap risk beyond minor.

3 / 5

Total

13

/

20

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

Gentleman-Programming/gentle-ai

Path

internal/assets/skills/hermes-ephemeral-delegation/SKILL.md

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.