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 skills/wu736139669/agent-taskgraph-protocol/agent-gep-litenpx skills add wu736139669/agent-taskgraph-protocol --skill agent-gep-litegit clone --depth 1 https://github.com/wu736139669/agent-taskgraph-protocolWrote 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/wu736139669/agent-taskgraph-protocol/agent-gep-lite)<a href="https://agentmods.dev/skills/wu736139669/agent-taskgraph-protocol/agent-gep-lite"><img src="https://agentmods.dev/badge/skills/wu736139669/agent-taskgraph-protocol/agent-gep-lite.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 | $0.00091 | $0.00940 |
| Opus 5 | $0.00046 | $0.00470 |
| Sonnet 5 | $0.00018 | $0.00188 |
| Haiku 4.5 | $0.00009 | $0.00094 |
Grade A, and why
agent-gep-lite scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Agent GEP-lite 使用指南 —— 经验层接入
你是 taskgraph 编排的经验扩展，不改变 PMO/Worker/Reviewer 的职责：只在两个点插一条线 —— 派发前召回、验收后回写。所有资产是纯本地文件（lessons/），经验库随仓库走（git 即同步）。
角色定位（与 taskgraph 的边界）
- PMO 的输入增强：召回结果只是 PMO 分诊/派发的决策输入（≤5 条、每条 ≤3 行），不是让 Worker 读一堆历史。
- Reviewer 的验收副产品：验收通过/失败都值得沉淀 —— 失败案例必须蒸馏成「避免:」短句，不堆长文（长文会稀释提示词）。
三个动作
1. 派发前召回（接第 7 节任务分配与派发）
L=plugins/agent-gep-lite/skills/agent-gep-lite/scripts/lesson-recall.sh
"$L" --bump --append queue/inbox/<task>.md "并行开发; 冲突面; 验收命令"
--append把「相关经验」段幂等写入简报（标记区间替换，重复跑不叠加）；- 召回时见 2+ 条同场景
failure→ 直接判"此块不可靠"，按第 7 节回退走单 agent 或澄清； - 命中高 score 同场景 Capsule → 简报里注明"参考经验：…"，Worker 可先读该胶囊原文再动手。
2. 验收后回写（接第 9 节验收（三级闸））
"$L" --kind success --gene gene-accept-run-command # 也可用 --help 看全参
以闸结果为准：通过 → kind: success；失败 → kind: failure（strategy 字段写「避免:」短句）。Reviewer 可以在验收汇报里附一句"已沉淀 capsule-xxxx"。
3. 台账维护
lessons/ledger.json是派生索引：lesson-record.sh --reindex重建；lesson-record.sh幂等（sha256 内容寻址），重复执行不产生重复条目；- 经验库随 git 仓库走；换机器 = clone 即可继承全部经验。
经验资产字段（详见 lessons/README.md）
| 字段 | 含义 |
|---|---|
signals |
触发场景标签（分号分隔），召回靠它匹配 |
strategy |
策略要点；失败案例以「避免:」开头 |
verify |
验收命令与关键输出（沉淀的依据） |
env |
环境指纹（repo/branch/os/date），判断经验是否适用 |
score |
0-1；失败案例低分，召回时权重自动降低 |
纪律
- 不把长文回执写进经验（教训 = 蒸馏 ≤3 行）；原文证据留在
queue/done/与 git diff 里； - 经验的质量取决于验收的严格程度 —— "说做完了"不算数（与 taskgraph 第 9 节一致）；
- 本插件不改 taskgraph 主流程：不开它，编排照常；开了它，只加"查"和"写"。
What ships with it
8 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.
- lessons/capsules/capsule-03c6cd42e12c.md 826 B
- lessons/capsules/capsule-625adb44da05.md 804 B
- lessons/genes/gene-accept-run-command.md 536 B
- lessons/genes/gene-conflict-map.md 640 B
- lessons/ledger.json 1.0 KB
- lessons/README.md 6.0 KB
- scripts/lesson-recall.sh 7.9 KB runs code
- scripts/lesson-record.sh 9.3 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.
- today First seen · 57 lines · 0 tokens per session scan A 0fec5fe0d81d
agent-gep-lite is a skill published in the GitHub repository wu736139669/agent-taskgraph-protocol (4 stars, last pushed yesterday), licensed Apache-2.0. It adds 91 tokens to every session and 940 once invoked, about $0.0005 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 2026-09-04.
Other skills, from other repositories
fabric-exec
Troubleshooting and advanced API reference for fabricexec TypeScript programs, dynamic providers, agents, and schema recovery. Routine pi. coding calls are documented by ambient guidance; load this skill only after an argument-shape error or when an advanced surface needs exact contracts.
fabric-fusion
Multi-model deliberation. Two to 8 distinct models answer in parallel with web-capable tools, then a judge compares consensus, contradictions, coverage gaps, unique insights, and blind spots. Act mode runs 1–4 read-only references, then one actor reconciles and executes. Use when the cost of being wrong justifies…
fabric-rlm
Recursively decomposes oversized tasks into bounded child Pi agents with fresh context windows. Use for whole-repo audits, massive-context analysis, and multi-file refactors that do not fit one context.
fabric-schema
Uses Fabric's typed Schema evidence loop and, when enabled, its bounded local-file transaction channel. Use when surprise must void a plan and mutation claims need explicit postconditions.
fabric-workflow
Runs a dynamic Pi Fabric workflow with code-held phases, fan-out, pipelines, structured agents, and best-effort verification. Use for large audits, migrations, parallel research, or explicit workflow requests.
fabric-council
Runs a bounded multi-perspective Pi Fabric council with independent reviewers and best-effort synthesis. Use for architecture choices, plans, reviews, and adversarial cross-checking.