diagnose-why-work-stopped

Diagnose stalled, looping, or over-recovered Paperclip issue trees and propose a no-code product-rule plan. Use when asked why work stopped, why it looped, or how to prevent a tree from going too deep.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

88%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 high-quality process skill: a clearly sequenced nine-step workflow with explicit gates, a pre-posting verification checklist, and unusually concrete operational details (state patterns, idempotency keys, assignee routing). The main weaknesses are mild redundancy around the execution-semantics reading instruction and no use of bundle reference files to slim the monolithic body.

Suggestions

State the 'read doc/execution-semantics.md' requirement once (e.g. in Step 0) and reference it from the other sections instead of repeating the instruction and its terminology list three times.

Move the catalog of common stop shapes (the five bullet patterns in Step 1) and/or the pitfalls into a `references/` file, keeping a one-line summary inline in SKILL.md, to shorten the main body and improve progressive disclosure.

Trim the quoted user commentary in Step 2 to its operative instruction ("survey what shipped in the last few days before proposing a rule").

Dimension	Reasoning	Score
Conciseness	The body is dense with non-obvious institutional specifics (issue-state combinations, idempotency key formats, invariant citations) and explains nothing Claude already knows. Not a 5: the instruction to read `doc/execution-semantics.md` is repeated three times (preamble, Step 0, Step 4), and Step 2 embeds quoted user commentary ("review our recent work on liveness that we shipped in the last couple of days.") that could be trimmed.	4 / 5
Actionability	Fully concrete, executable guidance for an instruction-only skill: exact stop-shape patterns ("`in_review` with no typed execution participant, no active run..."), an exact idempotency key format ("confirmation:{issueId}:plan:{revisionId}"), three explicit classification categories, and a phased plan structure with named assignee routing. Not a 4: no significant gaps — the guidance is copy-paste actionable within its domain.	5 / 5
Workflow Clarity	Nine explicitly numbered steps (0–8) in a coherent sequence, with explicit validation gates ("Do not propose a rule until you have a concrete stop point"; decompose only after acceptance) and a 9-item verification checklist before posting the plan. Error-recovery feedback loops are present ("If the rule would have blocked a recent productive run from succeeding, drop or narrow it"), which matters given Phase 0 mutates live issue trees.	5 / 5
Progressive Disclosure	Clear section headers make the ~150-line body easy to navigate, and the one external reference (`doc/execution-semantics.md`) is well-signaled and one level deep. Not a 5: the skill ships no bundle files at all — the stop-shape evidence catalog, pitfalls, and per-step detail are all inline in a single monolithic file where a reference file could offload them.	4 / 5
	Total	18 / 20 Passed

Description

82%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 strong description: it states a clear, distinctive capability with an explicit and well-phrased 'Use when' clause, and it stays concise in third-person voice without fluff. The only notable gap is trigger-term coverage — several natural synonyms that the body itself triggers on ("infinite loop", "stuck", "spinning", "root cause") are absent from the description.

Suggestions

Add the natural trigger synonyms the body already uses — e.g. "infinite loop", "stuck", "spinning", or "root cause" — to the 'Use when' clause to broaden trigger coverage.

Optionally name one more concrete capability (e.g. root-cause forensics on the tree, or producing a classification of every non-progressing issue) to lift the action coverage from two actions toward several.

Dimension	Reasoning	Score
Specificity	Names the domain ("Paperclip issue trees") and exactly two concrete actions ("Diagnose stalled, looping, or over-recovered... trees", "propose a no-code product-rule plan"), matching the anchor for domain plus 1-2 concrete actions that are not comprehensive. It is not a 4 because the description does not list several actions (forensics, classification, phased planning are all absent).	3 / 5
Completeness	Explicitly answers both what ("Diagnose stalled, looping, or over-recovered Paperclip issue trees and propose a no-code product-rule plan") and when ("Use when asked why work stopped, why it looped, or how to prevent a tree from going too deep") with concrete trigger phrases. Not a 4: the 'when' clause is explicit and specific, matching the top anchor.	5 / 5
Trigger Term Quality	Includes natural phrases a user would actually say: "why work stopped", "why it looped", "stalled", "looping", "going too deep". Not a 5 because common variations the skill body itself relies on — "infinite loop", "spinning", "stuck", "root cause" — are missing from the description.	4 / 5
Distinctiveness Conflict Risk	"Paperclip issue trees" and "over-recovered" carve out a clear niche with distinct triggers; this description would not plausibly fire for an unrelated skill. Uses third-person voice with no over-claims.	5 / 5
	Total	17 / 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
relative_links	Relative link issues: 19 suspicious	Warning

	Total	15 / 16 Passed

Repository: paperclipai/paperclip
Path: .agents/skills/diagnose-why-work-stopped/SKILL.md
Commit: 1c07b59

Reviewed: 1 day 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.