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/rivet-dev/agentos/update-acpnpx skills add rivet-dev/agentos --skill update-acpgit clone --depth 1 https://github.com/rivet-dev/agentosWhat 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 | $0.00052 | $0.00528 |
| Opus 5 | $0.00026 | $0.00264 |
| Sonnet 5 | $0.00010 | $0.00106 |
| Haiku 4.5 | $0.00005 | $0.00053 |
Grade A, and why
update-acp 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.
What it actually says
Update ACP
- Pin the latest
schema-v1.*ACP tag and the exact local harness and adapter versions. Keep the schema release, SDK/package release, and wireprotocolVersiondistinct. Ignore draft/v2 features unless clearly labeled. - Enumerate every stable ACP method, capability, content type, update, and lifecycle behavior. Include AgentOS-required extensions separately; do not present them as ACP requirements.
- Audit every agent registered under
registry/agent/. Prefer current upstream source over README claims:- OpenCode: native
opencode acpand its internal harness APIs. - Claude: Claude Agent SDK/CLI and
@agentclientprotocol/claude-agent-acp. - Codex: Codex App Server/CLI and
@agentclientprotocol/codex-acp. - Pi: Pi core, the pinned Pi RPC contract, and
svkozak/pi-acp.
- OpenCode: native
- Produce one table per agent. For every feature report: ACP requirement (
must, optional, or extension), harness support, harness control-interface support (RPC/App Server/SDK/CLI), upstream adapter support, AgentOS sidecar support, AgentOS public API/type support, confidence, and source evidence. Useyes,partial,no, orn/a. Group rows only when every status matches, and name every grouped feature. Name equivalent harness primitives such asswitch_session; do not require them to share the ACP method name, and never infer harness support only from adapter behavior. - Inspect AgentOS at minimum in
crates/agentos-sidecar/src/acp_extension.rs,packages/core/src/agent-session-types.ts,packages/core/src/agent-os.ts, registry manifests, and relevant tests. Verify both runtime forwarding and public exposure; preserved unknown JSON alone is not typed API support. - After the matrix, list prioritized action items grouped by owner: upstream adapter, harness/control interface, AgentOS sidecar/runtime, AgentOS API/types, or packaging/tests. State when no action is justified because the harness lacks the feature or ACP makes it optional.
- Do not change code during the audit. End by asking: Do you want me to fix these? If so, which priorities or agents should I include?
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 · 19 lines · 52 tokens per session scan A bc2e40634f43
update-acp is a skill published in the GitHub repository rivet-dev/agentos (4,444 stars, last pushed yesterday), licensed Apache-2.0. It adds 52 tokens to every session and 528 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 2026-08-30.
Other skills, from other repositories
maintain-model-list
Maintain the supported LLM model list: add a new model, or run routine maintenance to verify availability and discover new models worth adding. Use when the user asks to add/support a model, update the model list, or check model availability.
PuerTS Agent Development Guide
Guide for developing LLM agents based on PuerTsAgent framework — covers resource directory structure, system-prompt, skills, builtin modules, and best practices.
symbiont
AI-native agent runtime with typestate-enforced ORGA reasoning loop, Cedar policy authorization, CommunicationPolicyGate for inter-agent governance, ToolClad declarative tool contracts, knowledge bridge, zero-trust security, multi-tier sandboxing, webhook verification, markdown memory, skill scanning, metrics…
eve
Build durable backend AI agents with the eve framework. Use when creating, editing, or debugging an eve project — agent instructions, skills, tools, connections, channels, sandboxes, subagents, schedules, or evals.
skill-creator
创建新的 Claude Code 技能，修改和优化已有技能。当用户想从头创建技能、将当前工作流封装为技能、优化已有技能的内容或触发描述时使用此技能。即使用户没有明确说"技能"，当他们想把某个重复工作流程固定下来时也应使用。.
find-skills
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.