CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

self-evolving-single-agent

Use when generating a single installable agent that should keep learning, track sources, refresh research, propose repairs, or improve itself over time without becoming a multi-agent team.

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Medium

Suggest reviewing before use

Fix and improve this skill with Tessl

tessl review fix ./codex/plugins/agentlas-core-engine-meta-agent/runtime-support/skills/self-evolving-single-agent/SKILL.md

The canonical home for this skill is self-evolving-single-agent in agentlas-ai/Agentlas-OS

SKILL.md
Quality
Evals
Security

Quality

Content

71%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, well-structured instruction-only skill with a clear seven-step procedure and real approval-gate checkpoints. Its main weakness is undefined internal jargon and dangling references to contracts and mechanisms that are neither explained nor bundled, leaving the executor to guess their content.

Suggestions

Define or template the referenced mechanisms — e.g. briefly state what 'contracts/builder-interview-research-gate.md' must contain, or include it as a bundle file under references/.

Specify what the 'verification' output must check (e.g. validation steps for the generated package) to close the workflow's validation gap.

Clarify domain jargon like 'Memory Tickets', 'Memory Events', and 'revision-v1/evolution-v2 host capabilities' with a one-line definition each or a pointer to a reference file.

DimensionReasoningScore

Conciseness

The body is a lean numbered procedure with no explanation of concepts Claude already knows; only step 2's phrasing ('ask only material gaps, including zero questions when sufficient (8-12 only if that many gaps remain)') is convoluted and could be tightened.

4 / 5

Actionability

Concrete file manifests and named outputs give specific direction, but key details are missing: 'contracts/builder-interview-research-gate.md', 'Memory Tickets', and 'revision-v1/evolution-v2 host capabilities' are referenced with no definition, template, or accompanying bundle file.

3 / 5

Workflow Clarity

Steps 1-7 are clearly sequenced and the approval gates ('wait for human approval', 'one-use owner approval receipt', 'block activation on older hosts') plus the 'verification' output act as checkpoints, but validation of the generated package is named without ever being specified.

4 / 5

Progressive Disclosure

The skill is under 50 lines with no bundle files and well-organized sections (Procedure, Output, File revision approval); the referenced paths are artifacts of the generated package, not nested bundle references, so well-organized sections suffice per the simple-skill guideline.

5 / 5

Total

16

/

20

Passed

Description

50%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 a clear, distinctive niche and an explicit 'Use when' trigger, but it inverts the good pattern: it specifies when to use the skill without ever saying what the skill does. Trigger coverage is decent yet misses the most natural synonyms for the domain.

Suggestions

Add a 'what' clause stating the skill's operation, e.g. 'Generates the full package scaffolding — memory contract, research sources, approval gates, and global commands — for a self-improving single agent'.

Include the natural trigger synonyms users would say, such as 'self-improving', 'self-evolving', or 'autonomous agent', alongside the existing phrases.

Make the abstract capability verbs slightly more concrete (e.g. what 'refresh research' or 'propose repairs' produce) to strengthen specificity.

DimensionReasoningScore

Specificity

Lists several capability actions ('keep learning, track sources, refresh research, propose repairs, or improve itself') plus the concrete 'generating a single installable agent', but the actions remain abstract verbs without concrete mechanics, so coverage has minor gaps rather than being comprehensive.

4 / 5

Completeness

The 'when' is explicit ('Use when generating a single installable agent that should...'), but the description never states what the skill actually does — only 'when' is present without 'what', which is the score-2 anchor exactly.

2 / 5

Trigger Term Quality

Relevant phrases like 'single installable agent', 'improve itself', and 'multi-agent team' are present, but the most natural synonyms users would say for this niche — 'self-improving', 'self-evolving', 'autonomous agent' — are missing, matching the 'missing common variations or synonyms' anchor.

3 / 5

Distinctiveness Conflict Risk

'single installable agent... without becoming a multi-agent team' carves a clear niche with a distinguishing constraint, leaving only minor overlap risk with generic agent-generation skills rather than the minimal conflict of a fully distinct trigger set.

4 / 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

agentlas-ai/Agentlas-OS

Path

codex/plugins/agentlas-core-engine-meta-agent/runtime-support/skills/self-evolving-single-agent/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.