Agent Skills

ccf-literature-monitor

researchmikubaka881.4K installs

Scan recent arXiv/OpenReview papers, venue feeds, labs, and competitors for overlap. Use for 竞品监控, 新论文追踪, 论文跟踪, and recurring novelty watch. Give dated evidence and changes since the prior scan. Broad related-work retrieval belongs to ccf-literature-searcher.

Install

npx skills add https://github.com/mikubaka88/ccfa-skills --skill ccf-literature-monitor
SKILL.md

CCF Literature Monitor

Family File Contract

Before writing, resolve the canonical output and one stable working directory per task/artifact. Reuse explicit or established task paths; otherwise use project-root ccfa-workfiles/<purpose>/<artifact-id>/, with source/, assets/, cache/, and build/ only as needed. Update current files in place; do not scatter intermediates or create iteration copies. Preserve inputs and required evidence; clean only verified disposable files created by this task. Use UTF-8 text I/O and check Chinese text after saving or rendering. For file work, apply artifact-contracts.md and reuse the same paths across skill transitions.

Collaboration Contract

Before specialist execution, read and apply ccf-humanization first, then ccf-common. At every handoff, reuse their applicable active rules or refresh missing/changed ones. Both preflights are required even without prose; detailed editing, experiment, and maintenance modes run only when relevant.

Keep one integrating owner and actively use other skills to resolve missing prerequisites or check material findings. Reuse applicable evidence; do not skip necessary groundwork to save tokens. Before finalizing, integrate contributions and verify affected results. Follow the conditional cooperation routes; avoid unrelated stages and duplicate reports.

Core Rule

Monitor arXiv, OpenReview, conference/proceedings feeds, project pages, labs, and named competitors for new papers that could overlap with the user's idea or paper. Report findings factually. Do not exaggerate novelty threats, dismiss real overlap, or infer priority from weak evidence. Provide actionable signals: RELAX, RESEARCH, FOLLOW-UP.

Modes

  • arxiv-watch: scan recent papers in target categories or keyword clusters.
  • venue-watch: scan OpenReview, proceedings, accepted-paper lists, or official venue pages.
  • novelty-check: compare a given idea with recent papers and flag overlap.
  • trend-scouting: summarize emerging directions, crowded areas, and under-tested gaps.
  • competitor-tracking: monitor named research groups, labs, repositories, datasets, benchmarks, or authors.
  • paper-alert-digest: produce a recurring watch report suitable for saving into a project folder.

Skill Linkage

  • @READS ccfa.yaml for idea state.
  • @SHARES findings with ccf-literature-searcher.
  • @SHARES novelty implications with ccf-idea-reviewer.
  • @SHARES rescue or differentiation opportunities with ccf-idea-optimizer.
  • @SHARES citation and positioning candidates with ccf-paper-writer.
  • @FLAGS conflicting papers for ccf-integrity-auditor.

Invocation Controls

CCFA Handoff Mode: PARTIAL (Recommended). Follow metadata.ccf_skill_controls.handoff_question_mode and ../ccf-common/references/handoff-modes.md.

Load ../ccf-common/references/task-modes.md before deciding exploratory, quick, or standard mode.

Treat unpublished ideas, draft abstracts, methods, and results as private material. Load ../ccf-common/references/privacy-and-evidence.md before using private text in queries. Use public-safe keywords, public method names, venue names, and user-approved query text.

Load ../ccf-common/references/review-output-standards.md when monitor results feed idea scoring, paper review, or score-risk language.

Workflow

  1. Load references/monitoring-workflow.md.
  2. If project state is available (ccfa.yaml), read the current idea, target venue, claims, and tracked competitors.
  3. Choose the mode and time window. If the user says "latest", "recent", "new", "this week", or "today", verify the current date and use an explicit date range.
  4. Execute the requested monitoring mode using public-safe queries and source-quality exclusions from the shared policy.
  5. Classify overlaps by problem, mechanism, evidence, benchmark, dataset, claim, and venue positioning.
  6. Report new or materially changed findings, with precise implications for the requested research decision. Use existing report history to avoid presenting the same paper as new.
  7. Execute already requested downstream work through its owner; otherwise offer optional handoff to ccf-literature-searcher for deep retrieval, ccf-idea-reviewer for score impact, ccf-idea-optimizer for differentiation, or ccf-paper-writer for related-work integration.

Output Contract

For each monitoring execution, report in this structure:

## Literature Monitoring Report: [Mode] - [Topic/Venue]

**As of:** [YYYY-MM-DD]
**Time range:** [from] to [to]
**Sources scanned:** [arXiv categories / venues / labs / APIs]
**Papers scanned:** [count]
**High-relevance papers:** [count]
**Overall signal:** RELAX / RESEARCH / FOLLOW-UP

| Title | Source | Overlap level | Overlap type | Evidence basis | Action |
|:---|:---|:---:|:---|:---|:---:|
| [title] | [arXiv/OpenReview/venue] | None/Low/Medium/High | Problem/Method/Evidence/Benchmark | [abstract/intro/result] | RELAX/RESEARCH/FOLLOW-UP |

**Actionable flags:**
- RELAX: no material overlap found in the scanned window; continue, but do not claim full novelty proof.
- RESEARCH: partial overlap or possible close work; deep retrieve via `ccf-literature-searcher`.
- FOLLOW-UP: significant overlap in problem, mechanism, and evidence; route to `ccf-idea-reviewer` or `ccf-idea-optimizer` before writing stronger novelty claims.

**Handoff signals:**
- `ccf-literature-searcher`: [papers or clusters to retrieve deeply]
- `ccf-idea-reviewer`: [novelty or score implications]
- `ccf-idea-optimizer`: [differentiation or rescue direction]
- `ccf-paper-writer`: [related-work or positioning update]
- `ccf-integrity-auditor`: [citation/attribution conflict]

**Output self-check:** [date range explicit, source basis stated, table valid, no unsupported novelty conclusion]

References

  • references/monitoring-workflow.md — Workflow, api-key-management, search strategy.
  • references/report-template.md — Output template and formatting.
  • ../ccf-common/references/source-registry.yaml — Competitor and arxiv links.

Execution And Persistence

Distinguish a single scan from recurring monitoring. A recurring report format is not a scheduler: claim ongoing monitoring only after an actual authorized scheduled job exists. Record the time window and coverage; an empty scan is not proof of novelty. Deduplicate by stable paper identity and version, separating first publication from revisions. Batch independent feeds when supported; preserve real monitoring history under ../ccf-common/references/artifact-contracts.md. Follow no-new-files and existing project-state authorization.

Related skills

researchmattpocock575KInvestigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.paper-context-resolverlllllllama451KRigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing Renv-and-assets-bootstraplllllllama450KRigor Setup skill for README-first deep learning repo reproduction. Use when the task is specifically to prepare a conservative conda-first environment, checkpoint and dataset path assumptions, cache location hints, and setup notes before any run on a README-documented repository. Do not use for repo scanning, full orchestration, paper interpretation, final run reporting, or generic environment setup that is not tied to a specific reproduction target.ai-research-explorelllllllama311KRigor Explore compatible skill slug for meaningful and potentially novel deep learning research candidates. Use when the researcher has chosen the task family, dataset, benchmark, evaluation method, provided SOTA references, and wants candidate-only exploration on top of `current_research` with auditable repo understanding, idea gating, fair comparison, and governed experiments written to `explore_outputs/`. Do not use for README-first trusted reproduction, open-ended direction finding, narrow c

Search skills and MCP servers

Fuzzy search across 23,137 skills and servers