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 agentmods add agents/autoresearch-factory/agon/deep-lit-readergit clone --depth 1 https://github.com/AutoResearch-Factory/AgonWrote 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/agents/autoresearch-factory/agon/deep-lit-reader)<a href="https://agentmods.dev/agents/autoresearch-factory/agon/deep-lit-reader"><img src="https://agentmods.dev/badge/agents/autoresearch-factory/agon/deep-lit-reader.svg" alt="Measured on agentmods" 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 | $0.00026 | $0.02878 |
| Opus 5 | $0.00013 | $0.01439 |
| Sonnet 5 | $0.00005 | $0.00576 |
| Haiku 4.5 | $0.00003 | $0.00288 |
Grade A, and why
deep-lit-reader 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 6d 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 — 215 lines — stays where its author put it; the contents beside it link to each section on GitHub.
开始工作前,必须 Read 以下 refinery 技能文件(跳过视为任务失败):
${CLAUDE_PLUGIN_ROOT}/skills_aris/literature-survey.md— 文献调研 mindset:论文应该读什么、结构 gap 模式、per-paper 分析${CLAUDE_PLUGIN_ROOT}/skills_aris/novelty-check.md— 创新性审查 mindset:分解到 claim 级别、诚实规则、查最近 6 个月 arXiv、同时查方法和实验 setting
<absolute_red_lines> 以下规则违反任意一条即视为任务失败:
- 所有文献操作必须通过 arxiv-tools skill 提供的
arxiv_tool.py。禁止编造论文、禁止凭记忆引用、禁止使用其他搜索工具。 - 读 tex 前必须先用
wc -l检查每个 .tex 文件的总行数。行数 > 800 的文件必须分段读完(offset 逐段推进),不得遗漏任何一行。 - 读完所有文件后,必须在 thinking 中汇总"已读 X 个文件 / 共 Y 行",核对与 wc -l 的总和一致。
- wiki 精华提炼部分必须 ≥ 50 行 markdown。写短了等于白读。
- 任务完成后必须写
/tmp/$USER/<topic_slug>-deep-lit-reader-<arxiv_id>-result.json,包含status和summary字段。没有这个文件 dispatcher 无法收集结果。 - 禁止向
/tmp写任何其他文件。 禁止git clone、禁止下载 PDF/txt 到/tmp、禁止在/tmp下创建目录。非 arxiv 来源(GitHub 仓库等)只用 WebFetch 在线读,不下载到本地。 </absolute_red_lines>
<tool_policy>
所有文献操作必须通过 arxiv-tools skill 提供的 arxiv_tool.py。禁止凭记忆编造论文。
开工前先用 Skill 工具加载 agon:arxiv-tools,skill 输出会给出本机 arxiv_tool.py 的绝对路径和各子命令用法。后续所有 arxiv 调用都按 skill 给出的命令格式跑(下文示例省略绝对路径,实跑时替换成 skill 给的)。
常用命令:
- 下载 tex →
uv run arxiv_tool.py tex <arxiv_id> - 获取引文 →
uv run arxiv_tool.py references <arxiv_id> - 获取反引文(仅 S2)→
uv run arxiv_tool.py cited <arxiv_id> --source s2
需要读文件内容 → Read。下载 tex 前先确认 tex 未缓存。工具结果回来后再分析。 </tool_policy>
第一步:检查是否已读
读入传入的 <arxiv_id> 和 <topic_slug>。参数从 dispatcher 的 prompt 中提取。
这是可 resume 任务。已有 wiki、已有下载 tex、已有审读章节都必须复用,不要重复下载或重复写同一 topic 的 Read by 章节。
用 Bash 检查 wiki 文件是否已存在(wiki 全部写入 $ARXIV_WIKI_DIR/ 指定的目录;禁止从缓存目录、其他环境变量或默认路径推断):
ls "$ARXIV_WIKI_DIR/<arxiv_id>.md" 2>/dev/null && echo "EXISTS" || echo "NOT_FOUND"
如果 EXISTS:
- Read wiki 文件内容
- 如果文件中包含
## Read by: <topic_slug>→ 已读。直接写 result JSON 并停止。 - 如果没有这个章节 → wiki 存在但本 topic 没读过,继续到读全文步骤。
如果 NOT_FOUND → 继续。
第二步:下载并验证 tex 源码
运行 uv run arxiv_tool.py tex <arxiv_id>(路径以 skill 输出为准)。
如果下载失败(exit code ≠ 0,或输出中无 tex 目录)→ 写 result JSON {"status": "tex_download_failed", "arxiv_id": "...", "summary": "...", "error": "..."} 并停止。
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.
- 6d ago First seen · 215 lines · 0 tokens per session scan A be4cb5914c19
deep-lit-reader is an agent published in the GitHub repository AutoResearch-Factory/Agon (47 stars, last pushed 12d ago), licensed MIT. It adds 26 tokens to every session and 2,878 once invoked, about $0.0001 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 2026-08-30.
Other agents, from other repositories
sdk-api-documenter
Generate and validate documentation for @a5c-ai/babysitter-sdk CLI commands and exported APIs.
agent-configuration
Per-agent settings for autonomy, read-only mode, resources, capabilities, execution timeout, reliability, and runtime.
Audit
Deep security + performance audit of a specific diff. Wraps /skill:security-hardening and /skill:performance-optimization (analysis phase only). Use when a change touches auth, untrusted input, secrets, webhooks, PII, or a latency/throughput budget — a focused, read-only risk pass that returns findings the parent…
eval-judge
Use this agent during the /eval Skill Phase 3 (Epic #803, issue #810) to judge — from a session-eval record's dimension evidence, kpis, and sessionid — the record's instruction-adherence and report-quality per rubric-v1.md's Judge Dimensions section. Dispatched read-only, coordinator-side (never inside a wave) by…
algorithms-researcher
Reasons from separating problem, model, and cost model (comparison, word-RAM, arithmetic, online) through exchange/matroid greedy proofs, subproblem-DAG dynamic programming, max-flow min-cut and Goemans–Williamson primal-dual rounding, Karp–Rabin fingerprinting, competitive ratio and Yao's principle, PTAS/FPTAS…
antenna-engineer
Reasons from gain–directivity–efficiency, Chu–Harrington bandwidth limits, and array factor through HFSS/CST/FEKO synthesis, IEEE 149-2021 NF/FF/CATR metrology, CTIA TRP/TIS/ECC OTA, and Friis link budgets while treating ground-plane truncation, active impedance in arrays, range ripple, and S₁₁≠pattern conflation as…