Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add wanng-ide/open-paper-analysis --skill analyze-papergit clone --depth 1 https://github.com/wanng-ide/open-paper-analysisWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/wanng-ide/open-paper-analysis/analyze-paper)<a href="https://agentmods.dev/skills/wanng-ide/open-paper-analysis/analyze-paper"><img src="https://agentmods.dev/badge/skills/wanng-ide/open-paper-analysis/analyze-paper/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/wanng-ide/open-paper-analysis/analyze-paper"><img src="https://agentmods.dev/badge/skills/wanng-ide/open-paper-analysis/analyze-paper.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00116 | $0.01287 |
| Opus 5 | $0.00058 | $0.00643 |
| Sonnet 5 | $0.00023 | $0.00257 |
| Haiku 4.5 | $0.00012 | $0.00129 |
Grade A, and why
analyze-paper scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Analyze Paper
Analyze one research paper deeply and leave a verified, reusable note. Preserve the paper's actual mechanism and evidence instead of producing a generic summary.
Required references
Read these files before drafting:
- Configuration for target resolution, the version 2 schema, validation, and the credential boundary.
- Source policy for discovery and evidence rules.
- Content contract for the shared deep manuscript and multi-target invariants.
- Quality standard for the evidence map, chapter intent, figure/table analysis, and verification.
- Paper types after classifying the paper.
- Media policy before planning visual evidence.
Read the destination-specific reference before writing:
- Markdown output for the default output.
- Notion publishing only when the user asks for Notion or configured outputs include it.
- Lark publishing only when the user asks for Feishu/Lark or configured outputs include it.
Scope and defaults
- Handle one paper per run. If the request names multiple papers, ask the user to choose one; do not silently turn the task into a comparison or survey.
- Follow the user's language for the note. Preserve official names, metrics, benchmark names, and technical terms when translation would reduce precision.
- Resolve one or more destinations in this order: explicit user request, discovered configuration, Markdown.
- In a writable environment, write Markdown to
paper-notes/<paper-slug>.mdunless the user supplies a path. If file writing is unavailable, return the complete Markdown in the response. - Treat Notion as optional. Missing Notion tools or configuration must not block Markdown analysis.
- Treat Feishu/Lark as optional. Missing Lark tools or configuration must not block Markdown analysis.
- Default to deep analysis. Exhaust useful evidence without padding papers that do not support the full usual depth.
- Default media to numbered markers. Extract or upload figures only when the
user requests it or configuration sets media mode to
extract. - Treat each output independently. A remote publishing failure must not roll back another completed target.
What ships with it
12 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 270 B
- assets/config.example.toml 614 B
- references/configuration.md 2.9 KB
- references/content-contract.md 5.6 KB
- references/lark-publishing.md 4.8 KB
- references/markdown-output.md 5.7 KB
- references/media-policy.md 3.4 KB
- references/notion-publishing.md 6.1 KB
- references/paper-types.md 4.6 KB
- references/quality-standard.md 6.5 KB
- references/source-policy.md 3.5 KB
- scripts/validate_note.py 15 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 120 lines · 116 tokens per session scan A 03e0845103ec
analyze-paper is a skill published in the GitHub repository wanng-ide/open-paper-analysis (4 stars, last pushed 25d ago), licensed Apache-2.0. It adds 116 tokens to every session and 1,287 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
agency-os
Notion-as-source-of-truth dispatch board for running your work like an AI agency. One Tasks database is the source of truth; tasks flow Suggestion through Discussion, To-Do, In Progress, and Done with subtasks, recurring cadences, dependencies, and template subtrees. Batch execution fans approved To-Do rows out to…
live-dashboard
A Notion-style team dashboard template showing key performance indicators, seven-day trends, recent activity, and a task table. KPI means a measurable indicator of progress or performance.
notion-tool
Notion integration tool for searching, reading, creating, and updating pages and databases via the API. Use when: editing Notion pages, adding database rows, or searching a workspace.
deliverability
Use when the question is whether email will reach the inbox, covering sending domain setup, authentication, list hygiene, complaint and bounce rates, or a drop in delivery.
email-adaptation
Use when turning existing copy, a blog post, an announcement, or a brief into an email that works in an inbox, including deciding what to cut and what to link out to.
resend-build
Use when building, editing, or sending anything in Resend, including broadcasts, templates, segments, and contacts, or when checking what a past send did.