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/langerrr/zforge/retronpx skills add Langerrr/zforge --skill retrogit clone --depth 1 https://github.com/Langerrr/zforgeWrote 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/langerrr/zforge/retro)<a href="https://agentmods.dev/skills/langerrr/zforge/retro"><img src="https://agentmods.dev/badge/skills/langerrr/zforge/retro.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 | $0.00064 | $0.01400 |
| Opus 5 | $0.00032 | $0.00700 |
| Sonnet 5 | $0.00013 | $0.00280 |
| Haiku 4.5 | $0.00006 | $0.00140 |
Grade A, and why
retro 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 3d 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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Zforge Retro
Evaluate zforge's workflow performance on a feature. The output is for improving the zforge plugin itself — not for project documentation or tracking.
Run
- Convert the requested feature name to snake_case.
- Read
docs/{feature_name}/session_log.md. If it is missing or has no entries, stop with: "No session history to evaluate. Run the retro at the end of a working session." - Read
docs/{feature_name}/decision_review.md§B when present. These live process notes are the primary input. - Confirm this conversation contains substantial work on the feature. If this is a fresh session with no relevant implementation context, stop with: "Run the retro at the end of a working session, not in a fresh one."
- Create
docs/{feature_name}/.zforge-retro/and write the result to{session_id}.mdusing the output contract below. - Report the overall score, actionable findings, and any changes worth contributing to zforge.
Scope Boundary
Zforge is a workflow plugin, not a thinking framework. Retro judges:
- Templates, file structure, conventions
- Phase scaffolding, the report and recovery contracts
- Whether declared evidence was honoured, checked and carried
- Workflow ceremony and simplicity
Retro does NOT judge:
- Architectural or design decisions (superpowers territory)
- Code quality or correctness (review territory)
- LLM reasoning quality (model capability, not plugin)
- Whether the right feature was built (human's job)
The question is always: "Did zforge's workflow help or get in the way?"
NOT DO Rules
- Do not run in a fresh session. If the current session has no substantial conversation about the feature (no tool calls touching feature files, no implementation work), refuse. Retro requires session context to evaluate.
- Do not run if
session_log.mdhas no entries. No session history means nothing to evaluate. - Do not suggest zforge should make thinking decisions. If a finding starts drifting into "zforge should have caught this design flaw" — that's out of scope. Drop it.
- Do not produce findings without concrete evidence. Every finding needs specific files, sections, quotes, or examples. Vague observations are not findings.
- Do not aggregate across features. Each retro evaluates one feature's experience. Cross-feature analysis is done by humans through PRs to the zforge repo.
What ships with it
1 file 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.
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.
- 3d ago First seen · 119 lines · 64 tokens per session scan A 3d1978f0d50a
retro is a skill published in the GitHub repository Langerrr/zforge (10 stars, last pushed 3d ago), licensed MIT. It adds 64 tokens to every session and 1,400 once invoked, about $0.0003 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
math-unicode
Use whenever you need to express mathematical notation — equations, filters, set-builder, statistics, calculus, logic, ratios, drops, counts. Emit Unicode math glyphs INLINE; never wrap in $…$, \(...\), or $$...$$. Terminal coding agents (Claude Code, Codex CLI, and others) do not render LaTeX, so raw delimiters…
sdd-serve
Serve the SDD Builder's AI request queue: claim requests with sddnextrequest, draft the proposal, answer with sddrespondrequest. Never writes spec files — the user accepts each proposal in the builder. Use when the user asks to attend, serve or listen to the SDD board queue. / Atiende la cola de peticiones del SDD…
sdd-workflow
Guide a project with Spec-Driven Development (SDD) discipline - idea, approved spec, consistent plan, tasks, a gate that verifies approval and consent, implementation, validation, and logbook. Bilingual EN/ES. Use when the user wants to start, spec, plan, implement, or validate work with SDD, or mentions specs, plans…
eo-brainstorming
对不成形的想法做发散、对抗、拆解和方向决策。触发：帮我想想 / 头脑风暴 / brainstorming / /eo-brainstorming。 NOT FOR: 普通技术讨论或具体实现问题（只在用户明确"想想方向"时触发）。.
eo-change-review
对 change.md 做方案级审查（Delta 正确性、TODO 完整性、AC 覆盖）。触发：审查 change / change 审查 / 审方案 / /eo-change-review。 NOT FOR: 代码审查（/eo-review）、spec 审查（/eo-spec-review）、implement 内的回归审查。.
eo-miniapp-ideation
微信小程序需求挖掘与 MVP 构思，含变现模型和流量路径前置思考。触发：帮我想个小程序 / 小程序选品 / 小程序方向 / /eo-miniapp-ideation。 NOT FOR: 开发 / 备案 / 审核等执行流程。.