Tessl • The management layer for AI agent skills

Blog Docs Log inGet started

SkillPublic

continuous-agent-loop

Patterns for continuous autonomous agent loops with quality gates, evals, and recovery controls.

Quality

65%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/continuous-agent-loop/SKILL.md

Quality

Content

65%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 a tight, well-structured overview that routes to a production stack and catalogs failure modes and recovery moves, scoring well on conciseness and organization. Its weaknesses are missing execution details for the loop modes themselves and absent validation checkpoints in the recovery workflow.

Suggestions

Add a minimal invocation example or command for at least the default `sequential` loop mode so the reader knows how to start a loop, not just which one to pick.

Insert an explicit validation checkpoint in the recovery workflow (e.g. 'after `/harness-audit`, confirm the root cause is identified before reducing scope') and mark the recovery steps as an ordered sequence.

Move the 'v1.8+ canonical name / supersedes autonomous-loops' note into a short 'Deprecated aliases' section so version-sensitive details don't decay the main guidance.

Dimension	Reasoning	Score
Conciseness	The body is lean with no over-explanation of known concepts, but the opening 'This is the v1.8+ canonical loop skill name. It supersedes `autonomous-loops` while keeping compatibility for one release' is time-sensitive version information outside any 'old patterns'/'deprecated' section, incurring the rubric's minor penalty.	4 / 5
Actionability	There are some concrete anchors (a decision tree with explicit criteria, named tools like `ralphinho-rfc-pipeline`, `eval-harness`, `nanoclaw-repl`, and the `/harness-audit` command), but the skill never explains how to actually start or invoke any of the loop modes — the core execution detail is missing.	3 / 5
Workflow Clarity	Sequences are present (the selection flow and the recovery ordering of freeze → audit → reduce scope → replay), but validation checkpoints are absent or implicit, and loop automation is a batch-style repeated operation where the rubric caps workflow clarity at 3 without explicit validation steps.	3 / 5
Progressive Disclosure	The skill is under 50 lines, has no bundle files, and has no need for external references; its sections (Loop Selection Flow, Combined Pattern, Failure Modes, Recovery) are well-organized and easy to navigate, satisfying the rubric's simple-skill exception.	5 / 5
	Total	15 / 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 identifies a coherent niche and lists concrete topical elements, but it reads as a topic catalog rather than a capability statement and omits any 'when to use' trigger guidance. It is serviceable but sits at the midpoint of the rubric across all dimensions.

Suggestions

Add an explicit trigger clause, e.g. 'Use when running long-lived autonomous agent sessions, continuous iteration tasks, or when the user asks for a self-correcting loop with quality gates.'

Convert topic nouns into concrete actions, e.g. 'Selects, runs, and supervises continuous agent loops with quality gates, eval harnesses, and failure recovery.'

Include natural synonyms users would actually say ('keep going autonomously', 'iterate until passing', 'agent loop') to improve trigger-term coverage and distinctiveness from sibling loop skills.

Dimension	Reasoning	Score
Specificity	The description names the domain ('continuous autonomous agent loops') and concrete elements ('quality gates, evals, and recovery controls'), but these are topic nouns rather than actions — no verbs describe what the skill actually does with them, matching the anchor for domain plus some concrete but non-comprehensive content.	3 / 5
Completeness	The 'what' is reasonably clear (patterns for continuous agent loops with gates, evals, and recovery), but the 'when' is entirely absent — there is no 'Use when...' clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Terms like 'agent loops', 'evals', 'quality gates', and 'recovery controls' are relevant to the domain, but common natural variations a user would say ('run continuously', 'iterate', 'self-correcting loop', 'autonomous iteration') are missing.	3 / 5
Distinctiveness Conflict Risk	'Continuous autonomous agent loops with quality gates, evals, and recovery controls' is somewhat specific, but it could still overlap with other automation/loop skills — the body itself acknowledges sibling loop skills (continuous-pr, rfc-dag, infinite, sequential), indicating real overlap risk within its own ecosystem.	3 / 5
	Total	12 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: valeriibeck-star/claude-skills
Path: skills/continuous-agent-loop/SKILL.md
Revision: 0bda8b8

Reviewed: 2 days ago

Discovery Implementation Validation

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.

continuous-agent-loop

Table of Contents

Is this your skill?