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 well-structured, highly actionable diagnostic skill body with a clear sequenced procedure, strong validation checkpoints and feedback loops, and a useful verification checklist. It is slightly verbose in places and could offload some procedural detail to a reference file.

Suggestions

Move the per-phase assignee roster (CodexCoder/ClaudeCoder/UXDesigner/SecurityEngineer/QA) and the detailed phase decomposition into a references/plan-template.md file, leaving SKILL.md to summarize the workflow and signal the reference.

De-duplicate the execution-semantics terms (listed in both the intro and step 0) and the three invariants (intro and again in step 4) by referencing the earlier statement instead of restating it.

Consider extracting the common stop-point shapes ('in_review with no participant', 'in_progress after successful run', etc.) into a references/stop-shapes.md checklist so the core procedure stays scannable.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence (no 'what is an issue tree' padding), with load-bearing detail like invariants, forensics shapes, and pitfalls earning their tokens; minor repetition (execution-semantics terms listed in both the intro and step 0, invariants restated in step 4) keeps it just below a 5.	4 / 5
Actionability	As an instruction-only diagnostic/planning skill it provides concrete, executable guidance — exact issue-state shapes to look for, an idempotency key format ('confirmation:{issueId}:plan:{revisionId}'), named per-phase assignees, and blockedByIssueIds wiring — so the absence of code is not penalized.	5 / 5
Workflow Clarity	A clearly sequenced 9-stage procedure (step 0 through 8) with explicit validation checkpoints (forensics before proposing, recent-work survey before a new rule, approval gate before decomposition), feedback loops (revision supersedes prior confirmation), and a final verification checklist.	5 / 5
Progressive Disclosure	Clean, header-organized single file with one clearly signaled one-level reference (doc/execution-semantics.md) and no nested chains, but the procedure is long for a SKILL.md overview and some step detail (e.g. per-phase assignee rosters) could plausibly live in a reference file.	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 that clearly states a specialized domain, gives an explicit 'Use when' trigger clause with natural phrasing, and is well-differentiated from other skills. The only gap is that it lists just two concrete actions, capping specificity below the top level.

Dimension	Reasoning	Score
Specificity	Names the domain ('stalled, looping, or over-recovered Paperclip issue trees') and two concrete actions ('Diagnose...and propose a no-code product-rule plan'), matching the anchor for 1-2 concrete actions rather than the several required for a 4.	3 / 5
Completeness	It explicitly answers both 'what' (diagnose issue trees and propose a no-code product-rule plan) and 'when' with concrete trigger phrases via the 'Use when asked...' clause, matching the top anchor.	5 / 5
Trigger Term Quality	'Use when asked why work stopped, why it looped, or how to prevent a tree from going too deep' captures natural phrases a user/manager would say, with synonym coverage (stalled, looping, going too deep), but misses a few common variants like 'spinning' or 'stuck', keeping it just below a 5.	4 / 5
Distinctiveness Conflict Risk	The narrow niche ('Paperclip issue trees', 'over-recovered', 'no-code product-rule plan') and distinctive trigger phrasing make it clearly separable from other skills with minimal conflict risk.	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: bd1fdc2

Reviewed: 5 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.