research-pipeline

Full end-to-end research pipeline: from a broad research direction through idea discovery, experiments, and review all the way to a polished paper PDF. Use when user says "全流程", "full pipeline", "从找idea到投稿", "end-to-end research", or wants the complete autonomous research lifecycle.

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Quality

Content

73%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 pipeline skill with explicit validation checkpoints and resume logic. Its main weakness is redundancy: the overnight-heartbeat/stall-detection content is stated twice, which both pads the body and blurs progressive disclosure.

Suggestions

Collapse the opening heartbeat note and the "Overnight heartbeat" section into one place, keeping the detailed helper-resolution chain and pivot rules only in the dedicated section (or delegated to external-cadence.md) to remove the duplication.

Replace remaining $VARIABLE placeholders in example invocations (e.g. "$CHOSEN_IDEA_TITLE") with short inline examples or a named constants reference so the commands are copy-paste ready.

Move the full stall-detection bash block and `pivot` action table into shared-references/external-cadence.md and keep only a one-line pointer + the warn-and-skip rule in SKILL.md, improving both conciseness and progressive disclosure.

Dimension	Reasoning	Score
Conciseness	Mostly operational and assumes Claude's intelligence (no basic-concept padding), but the stall-detection/heartbeat material is duplicated between the opening note and the "Overnight heartbeat" section, repeating the helper-resolution chain and pivot rules.	3 / 5
Actionability	Concrete, executable guidance throughout (skill invocations, the bash helper-resolution chain, `python3 "$ITER_LOG" note ...`, `run_state.py` subcommands with args), with minor gaps where `$VARIABLE` placeholders remain.	4 / 5
Workflow Clarity	Clear Stage 1→5 sequence with Gate 1/Gate 2, a per-phase acceptance-gate table, resume re-validation of done-but-unaccepted stages, and explicit fail-gracefully / round-4-stop feedback loops for an expensive autonomous operation.	5 / 5
Progressive Disclosure	No bundle files present; references are one level deep and clearly signaled (shared-references/*.md, templates/RESEARCH_BRIEF_TEMPLATE.md). The inlined, duplicated heartbeat detail could instead live entirely in external-cadence.md.	4 / 5
	Total	16 / 20 Passed

Description

95%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, third-person description that clearly states the skill's full-pipeline purpose and gives explicit, natural (bilingual) trigger phrases. Stage-level action descriptions are the only minor weakness versus fully granular operations.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete pipeline stages ("idea discovery, experiments, and review all the way to a polished paper PDF"), but the actions are stage-level rather than fine-grained operations, leaving minor coverage gaps.	4 / 5
Completeness	Explicitly answers both what ("Full end-to-end research pipeline: from... to a polished paper PDF") and when ("Use when user says...") with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive natural trigger coverage including bilingual synonyms ("全流程", "full pipeline", "从找idea到投稿", "end-to-end research") that users would actually say.	5 / 5
Distinctiveness Conflict Risk	Clear orchestrator niche with distinct full-pipeline triggers; minimal overlap risk with the individual sub-skills it chains.	5 / 5
	Total	19 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
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
relative_links	Relative link issues: 6 suspicious	Warning

	Total	13 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/research-pipeline/SKILL.md
Commit: f4f20f9

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.