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/yimwoo/hotl-plugin/governed-executionnpx skills add yimwoo/hotl-plugin --skill governed-executiongit clone --depth 1 https://github.com/yimwoo/hotl-pluginWrote 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/yimwoo/hotl-plugin/governed-execution)<a href="https://agentmods.dev/skills/yimwoo/hotl-plugin/governed-execution"><img src="https://agentmods.dev/badge/skills/yimwoo/hotl-plugin/governed-execution.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 | $0.00023 | $0.00889 |
| Opus 5 | $0.00012 | $0.00445 |
| Sonnet 5 | $0.00005 | $0.00178 |
| Haiku 4.5 | $0.00002 | $0.00089 |
Grade A, and why
governed-execution 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 — 43 lines — stays where its author put it; the contents beside it link to each section on GitHub.
HOTL Governed Execution
This is the preferred execution entry point for new HOTL runs. It selects a host driver and an execution profile while preserving the canonical state machine in loop-execution.
Required behavior
- Resolve and lint the workflow using
loop-executionrules. - Select the profile from user intent:
loopfor autonomous sequential work,manualfor explicit checkpoints, ordelegatedfor eligible independent steps. Never select delegated execution when the user prohibited subagents. - Locate the router in the active HOTL installation, then run
runtime/drivers/route.sh preflight <workflow>before initialization. Supported roots include the current repository,~/.codex/hotl,~/.codex/plugins/hotl-source, the active Codex plugin cache, the Claude plugin root, and~/.cline/hotl. Use--hostonly when the user or trusted host context identifies one. - If preflight resolves fallback, follow the chosen existing execution skill with the generic driver. If it resolves native, obtain
envelopeand use only supported native features; the host sandbox and approvals remain authoritative. Native goals, automations, background sessions, handoffs, and hooks provide scheduling and liveness only. They never replace HOTL ownership, verification, state, or receipts. Preview and experimental continuation features remain explicit opt-ins. - Launch new driver-managed runs with
--require-owner. Immediately runowner claim --owner <stable-controller-id> --lease-seconds <bounded-lease> --run-id <run-id>, retain the returned token only in the controller, export it asHOTL_OWNER_TOKEN, and runowner heartbeatbefore and after long actions and at safe transition boundaries. Every later mutation must carry that token. Use explicitowner handoff,owner release, or reviewedowner takeover; never infer ownership from age alone. - Persist every step, verification, gate, action decision, effect outcome, budget observation, finalize, and finish transition through the selected driver. Host UI or chat text is never state.
- Before claiming success, require
receipt <run-id>to returnsufficiency.sufficient: true. A successfulfinalizeonly moves the run toready_to_finish;finishrecords the explicit disposition and moves it tocompleted. If interrupted, runreconcile <run-id>and followresumingverify-first behavior.
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 · 43 lines · 23 tokens per session scan A 225fda9d67a0
governed-execution is a skill published in the GitHub repository yimwoo/hotl-plugin (26 stars, last pushed 2mo ago), licensed MIT. It adds 23 tokens to every session and 889 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 skills, from other repositories
first-plan-provenance-tracker
Skill que define o schema de proveniência e o sistema de freshness tracking para findings em .first-plan/. Use durante Discovery e refresh para registrar source (file:line@SHA), TTL, confidence inicial e cadeia de supersedeção. Aplica decay de confidence baseado em idade.
first-plan-lens-engine
Skill que define o motor de detecção e roteamento de stack lenses do first-plan. Mapeia manifestos (go.mod, package.json, etc) -> stack -> papel -> lens skill correspondente. Use quando precisar decidir qual lens aplicar a uma pasta durante Discovery, ou quando estender suporte a uma nova stack.
first-plan-pattern-extraction
Skill que define como extrair padrões do código com confidence scoring. Use durante Discovery quando precisar identificar convenções, anti-padrões ou padrões idiomáticos do projeto. Cada padrão extraído deve vir com exemplo concreto do código (path:line).
first-plan-reuse-indexing
Skill que define como construir o Reuse Index invertido em 03-reuse/. Use durante Discovery para identificar componentes, utils, types, hooks reutilizáveis no projeto e criar o índice "preciso de X -> use Y".
first-plan-lens-terraform
Stack lens para Terraform e IaC similar (OpenTofu, Pulumi). Use durante Discovery quando arquivos .tf forem detectados. Cobre estrutura de módulos, backend de state, environments, providers, naming/tagging.
first-plan-lens-go
Stack lens para projetos Go. Use durante Discovery quando go.mod for detectado. Extrai padrões de layout (cmd/internal/pkg), error handling, context.Context, concurrency, code generation, build tags.