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/fold-run/fold/roadmapnpx skills add fold-run/fold --skill roadmapgit clone --depth 1 https://github.com/fold-run/foldWhat 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 | $0.00075 | $0.01122 |
| Opus 5 | $0.00037 | $0.00561 |
| Sonnet 5 | $0.00015 | $0.00224 |
| Haiku 4.5 | $0.00007 | $0.00112 |
Grade A, and why
roadmap 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 2d 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 — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The gap list and the roadmap
These are not marketing documents. gateway/era_test.go and
gateway/listen_test.go fail with messages that point the reader at
them, and CLAUDE.md's fourth rule makes updating them a merge condition.
A gap here is load-bearing text.
Two surfaces, different jobs:
- README "Not implemented" — what fold does not do today, with the reason. Read by someone evaluating fold, so each entry has to survive a skeptic: name the gap, name the cause (structural, waiting on someone else, or deliberate), and say what would close it.
- docs/roadmap.md — what fold intends, in three horizons plus "Non-goals" and "Deprecations on the clock". No dates, deliberately.
Every "Not implemented" entry has a roadmap counterpart. The README says what is missing; the roadmap says whether it is coming.
The four vetoes
docs/roadmap.md "How to read this" names what kills an item regardless of
demand:
- the v1 compatibility contract and the frozen defaults,
- the invisibility rule,
- the added-latency gate,
- audit as the single exit door.
A feature that requires breaking one does not get built — it gets a non-goal entry with the reasoning written down. That conversion is the main judgement this skill exists for. Content inspection is the worked example: it is a non-goal because inspecting bodies means buffering and rewriting, which conflicts with (2) and (3) — and the entry still offers the fold-shaped answer (the decision hook) rather than stopping at "no".
A non-goal entry that only says no is incomplete. Say what the veto was and what someone with that need should do instead.
Closing a gap
When a change closes one, the same PR does all of it:
- Remove or narrow the README "Not implemented" entry — narrowing is common and honest: three consequences, one addressed, two named.
- Update the roadmap counterpart (move it out of its horizon, or strike it).
- Find the canary. If the gap was waiting on someone else, there is probably a test that fails when the wait ends. Flip or delete it, and never "fix" one by loosening the assertion.
- CHANGELOG entry if it ships in a release.
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.
- 2d ago First seen · 107 lines · 75 tokens per session scan A e2c036d26d8e
roadmap is a skill published in the GitHub repository fold-run/fold (1 stars, last pushed 2d ago), licensed Apache-2.0. It adds 75 tokens to every session and 1,122 once invoked, about $0.0004 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
hotplex-release
通过判断 SemVer 影响、整理 CHANGELOG、对账版本面和验证发布输入来准备或发布 HotPlex 版本。不要用于主机二进制更新、运行时诊断或文档巡逻。.
hotplex-docs-patrol
维护 HotPlex 当前文档，将代码、配置、API 或发布变更映射到 BFS 可达文档，只修复已验证的文档漂移。不要用于普通润色、运行时诊断或版本发布。.
hotplex-stt-tts
初始化或修复 HotPlex 本地 STT 和 MOSS TTS 运行时，包括 Python 依赖、官方模型、配置和验收。仅在明确授权主机变更时使用；只读检查请使用 hotplex-cli 或 hotplex-diagnostics。.
hotplex-diagnostics
深入诊断 HotPlex Gateway、Worker、Session、日志或反馈链异常。普通 status/doctor/security/config 只读检查属于 hotplex-cli；安装、更新、重启、配置写入和 Admin 变更属于 hotplex-operator。.
hotplex-cli
使用 HotPlex CLI 处理 Cron、明确请求的 Slack 操作、普通用户聊天命令指引，以及只读 status、doctor、security、config 诊断。不要用于飞书写操作、发布、服务安装、二进制更新或 Admin 变更。.
hotplex-operator
运维或初始化 HotPlex 主机，覆盖首次 onboard、服务安装/启动、二进制更新、主机配置、审计检查和 Admin 变更。仅在明确授权的 operator 上下文中使用。.