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/preflightnpx skills add fold-run/fold --skill preflightgit 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 | $0.00070 | $0.00912 |
| Opus 5 | $0.00035 | $0.00456 |
| Sonnet 5 | $0.00014 | $0.00182 |
| Haiku 4.5 | $0.00007 | $0.00091 |
Grade A, and why
preflight 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
fold preflight
CI runs eight jobs: test (gofmt, go mod tidy -diff, vet, build,
go test -race + coverage), lint, vuln, bench, helm,
conformance, console, and image. make check covers test and
lint; everything else is conditional below. This skill runs the local
equivalent, cheapest-first, and stops early on failure.
Always
make check # fmt-check + tidy-check + vet + build + race + lint
Run failures to ground before moving on:
- fmt-check →
make fmt, re-run. - tidy-check →
go mod tidy, then review the go.mod/go.sum diff — an unexpected dependency change is a finding, not a fix. - race failures → these are real bugs in a gateway; never "fix" by
removing
-raceor adding sleeps. Suspect snapshot access outside the atomic load, session-map access outside its lock, or test fixtures sharing state. - lint → config is
.golangci.yml; fix the code, don't add nolint without saying so.
Conditional gates — run when the diff touches them
- Proxy path (
gateway/router.go,upstream.go,endpoints.go, middleware, egress rewriting) →make bench. Gate: added p50 < 5 ms through the proxy path. A regression means a hot-path allocation or synchronous work that belongs at snapshot-build time. - Behavior through the gateway (anything a client could observe) →
make conformance. Needs node/npx; expects 40/40. A failed check usually means the gateway stopped being invisible — compare against hitting the upstream directly before touching the pinned commit. deploy/helm/→make helm-check. Lint, render against everyci/*.yaml, and the required-config guard.go.mod/go.sum/ a Go toolchain bump →make vuln. This is its own CI job, not part ofmake check, so it is easy to skip and easy to be surprised by. A red govulncheck is often the live advisory database rather than the diff — check the Go patch version ingo.modagainst the advisory before hunting for a bug in the change.gateway/console/,scripts/sync-console.sh, or aCONSOLE_COMMITbump →make console-check(needs network). Proves the vendored bytes are exactly the pinned fold-console commit. A hand edit undergateway/console/cannot merge at all — the PreToolUse guard blocks it, and CI'sconsolejob would fail it. See/console-sync.deploy/docker/*.Dockerfile,go.mod's Go version, or anything the images build from → build the images locally:
CI'sdocker build -f deploy/docker/fold.Dockerfile --build-arg VERSION=ci .imagejob covers this per-PR now, but it exists because a builder base pinned below the versiongo.modrequires sat on main through two green merges and failed after the tag was pushed — with the binaries and chart already published.- Untrusted-input parsers (config parse, cursors, discovery docs) →
make fuzzfor a deeper pass; seed corpora already run inmake test.
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 · 70 lines · 70 tokens per session scan A 3705438d56fa
preflight is a skill published in the GitHub repository fold-run/fold (1 stars, last pushed 2d ago), licensed Apache-2.0. It adds 70 tokens to every session and 912 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
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。.