paper-writing

Workflow 3: Full paper writing pipeline. Orchestrates paper-plan → paper-figure → figure-spec/paper-illustration/mermaid-diagram → paper-write → paper-compile → auto-paper-improvement-loop to go from a narrative report to a polished PDF. At `— effort: max | beast` (or explicit `— assurance: submission`), Phase 6 gates the Final Report on `verify_paper_audits.sh` (resolved per integration-contract §2); the PDF is labelled `submission-ready` only when the external verifier is green. Use when user says "写论文全流程", "write paper pipeline", "从报告到PDF", "paper writing", or wants the complete paper generation workflow.

Quality

77%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/skills-codex/paper-writing/SKILL.md

Quality

Discovery

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

This is a strong skill description that clearly articulates a complex multi-step pipeline, provides explicit trigger terms in multiple languages, and distinguishes itself well from component sub-skills. The description is detailed and specific about both capabilities and activation conditions. Minor concern is that the internal jargon (e.g., 'integration-contract §2', effort flags like '— effort: max | beast') adds complexity that may not aid skill selection, but overall it is comprehensive and well-constructed.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions in a pipeline: paper-plan, paper-figure, figure-spec, paper-illustration, mermaid-diagram, paper-write, paper-compile, auto-paper-improvement-loop. Also describes specific gating behavior with verify_paper_audits.sh and submission-ready labeling.	3 / 3
Completeness	Clearly answers both 'what' (orchestrates a full pipeline from narrative report to polished PDF with specific phases and quality gating) and 'when' (explicit 'Use when...' clause with specific trigger phrases and a general condition).	3 / 3
Trigger Term Quality	Includes natural trigger terms in both English and Chinese: '写论文全流程', 'write paper pipeline', '从报告到PDF', 'paper writing', 'complete paper generation workflow'. These cover natural user phrasings well, including bilingual variations.	3 / 3
Distinctiveness Conflict Risk	Highly distinctive as a full paper writing pipeline orchestrator. The specific pipeline steps, the mention of verify_paper_audits.sh, submission-ready gating, and bilingual triggers make it clearly distinguishable from individual sub-skills like paper-write or paper-compile alone.	3 / 3
	Total	12 / 12 Passed

Implementation

55%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This skill excels at actionability and workflow clarity — every phase has concrete commands, explicit validation gates, and clear error recovery paths. However, it is severely over-long and monolithic, with Phase 6 alone containing hundreds of lines of inline bash scripts, checklists, and templates that should be in referenced files. The skill reads more like a complete specification document than a concise orchestration guide, wasting significant context window budget on content that Claude could reference from separate files.

Suggestions

Extract Phase 6's verifier resolution logic, submission pre-flight checklist, and final report template into separate referenced files (e.g., SUBMISSION_GATE.md, REPORT_TEMPLATE.md) to reduce inline bulk by ~40%.

Remove 'Empirical motivation' paragraphs (Phases 5.5 and 5.8) — these explain why the skill exists rather than instructing Claude what to do, and Claude doesn't need persuasion to follow instructions.

Consolidate the Phase 2b illustration mode descriptions into a compact table rather than four separate verbose blocks with repeated output/best-for descriptions.

Move the 'Optional hardening' and 'Composing with Other Workflows' sections to a separate ADVANCED.md reference file, as they are supplementary context rather than core pipeline instructions.

Dimension	Reasoning	Score
Conciseness	This skill is extremely verbose at ~500+ lines. It contains extensive explanations of assurance levels, verifier resolution chains, escape hatches, empirical motivations, and optional hardening sections that could be dramatically condensed. Phase 6 alone is longer than many entire skills. Much of this content (e.g., explaining what each audit does, empirical motivation paragraphs, the full verifier resolution bash script) could be in referenced files rather than inline.	1 / 3
Actionability	The skill provides fully concrete, executable guidance throughout: specific bash commands, exact invocation syntax for each sub-skill, complete bash scripts for verifier resolution, regex patterns for claim detection, and precise checkpoint templates. Every phase has copy-paste ready commands and clear expected outputs.	3 / 3
Workflow Clarity	The multi-step pipeline is exceptionally well-sequenced with explicit phases (0-6), clear checkpoints between each phase with specific output templates, validation gates (proof verification, claim audit, citation audit, external verifier), and explicit feedback loops (fix → re-validate → proceed). The submission gate in Phase 6 has thorough error recovery guidance for each failure mode (MISSING, STALE, BLOCKING_VERDICT, SCHEMA_INVALID).	3 / 3
Progressive Disclosure	Despite referencing external files (integration-contract.md, assurance-contract.md, reviewer-independence.md, output protocols), the skill is a monolithic wall of text with massive inline content that should be split into separate files. Phase 6 alone contains verifier resolution scripts, optional hardening configs, pre-flight checklists, and the full final report template — all inline. No bundle files are provided to support the references made, and the content that is inline far exceeds what belongs in an overview skill.	1 / 3
	Total	8 / 12 Passed

Validation

72%

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 — 8 / 11 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (633 lines); consider splitting into references/ and linking	Warning
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	8 / 11 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Commit: a425a71

Reviewed: about 23 hours 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.