continue

Use after an execution slice to reassess whether the current tactic still serves the user's north star, then keep course, pivot, replan, or stop. Designed for long-running loops where the agent should keep moving unless a real blocker or completed lane exists.

Quality

74%

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 ./.agents/skills/continue/SKILL.md

Quality

Content

81%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 highly actionable, well-sequenced instruction skill: every verdict, gate, and output field is concrete and immediately executable, with a clear decision workflow and validation checkpoints. Its weaknesses are redundancy across sections (state-field lists and `blocked` rules repeated) and the lack of any progressive disclosure despite the length.

Suggestions

Deduplicate the state-file field lists ("Mutable State Rule" vs "State Update Checklist") and state the `blocked` semantics once — currently repeated in Completion Guard and the Stop section.

Move the Escalation Rules detail or the State Update Checklist into a reference file (e.g. references/escalation.md) and keep a one-line pointer in SKILL.md to reduce the always-loaded token footprint.

Dimension	Reasoning	Score
Conciseness	The body is terse and imperative with zero conceptual padding, but carries real redundancy: the state-file field list appears in both "Mutable State Rule" and "State Update Checklist", and the `blocked`-semantics rules are stated three times (Completion Guard, the Stop verdict, and the pre-stop checklist). This matches the anchor for mostly efficient content that could be tightened, rather than the minor-trim bar for 4.	3 / 5
Actionability	Fully concrete guidance throughout: exact verdict names, a copy-paste output template (`Verdict:` / `Harsh take:` / `Why:` / `Risks:` / `Earliest gates:` / `Next move:` / `Do not do:`), exact state-file fields and status values (`pending`, `in_progress`, `revise`, `done`, `blocked`), a mechanical decision rule ("If the answer to 2, 3, 4, or 5 is yes, pivot or replan"), and a full worked example. As an instruction-only skill the absence of code is not penalized, and the guidance is fully executable.	5 / 5
Workflow Clarity	The multi-step process is clearly sequenced (read the narrowest state file → answer Core Questions → choose a verdict → execute the next move → update state in the same turn) with explicit validation checkpoints: the Completion Guard, earliest gates, verdict preconditions, and a keep-moving anti-stall loop. Checkpoints are explicit and complete, matching the 5-anchor rather than the minor-gaps 4-anchor.	5 / 5
Progressive Disclosure	No bundle files exist and the body is a single well-organized file with clear, navigable sections, all of which apply at decision time — good structure with content mostly appropriately placed. It falls short of 5 because at ~390 lines with no references, content such as the escalation rules or the state update checklist could be split into a one-level-deep reference file, and the under-50-line exception does not apply.	4 / 5
	Total	17 / 20 Passed

Description

67%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.

A solid description that explicitly covers both what the skill does and when to use it, with a distinctive verdict vocabulary. Its main weakness is trigger phrasing built on internal jargon ("execution slice", "north star", "owner") rather than words a user would naturally say when needing this skill.

Suggestions

Rewrite the trigger in the user's vocabulary, e.g. "Use when a long task is in progress and you must decide whether to continue, change approach, replan, or stop" — replacing jargon like "execution slice" and "north star" with natural phrases.

Add common trigger variations and synonyms ("keep going", "should I continue", "mid-loop checkpoint", "autonomous loop") to improve keyword coverage for natural invocation.

Name the concrete artifacts the skill touches (e.g. the mutable state file, the completion state file) so the what-clause is grounded in specific objects, not only abstract actions.

Dimension	Reasoning	Score
Specificity	The description names the domain (progress reassessment in long-running loops) and lists several concrete actions — "reassess whether the current tactic still serves the user's north star, then keep course, pivot, replan, or stop" — matching the anchor for several specific actions with minor gaps. It falls short of 5 because the actions stay abstract rather than naming concrete artifacts, commands, or file types.	4 / 5
Completeness	Both what ("reassess whether the current tactic still serves the user's north star, then keep course, pivot, replan, or stop") and when ("Use after an execution slice" plus "Designed for long-running loops") are explicitly present. Not 5 because the trigger is framed in domain jargon rather than concrete natural trigger phrases.	4 / 5
Trigger Term Quality	Terms like "long-running loops", "keep moving", and "blocker" are relevant, but the natural phrases a user would actually say (e.g. "keep going", "when to stop", "should I continue") are missing, and "execution slice" / "north star" are internal jargon rather than user vocabulary. This matches the anchor for some relevant keywords but missing common variations.	3 / 5
Distinctiveness Conflict Risk	The progress-governor framing with distinct verdict vocabulary (keep course / pivot / replan / stop) carves a clear niche with minimal overlap. Minor overlap risk remains with generic loop, planning, or autonomy-continuation skills, so it is mostly distinct rather than fully 5.	4 / 5
	Total	15 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: udecode/plate
Path: .agents/skills/continue/SKILL.md
Commit: babb3c2

Reviewed: 2 days ago

Table of Contents

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.