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 agents/rimio-ai/rimz/adapter_opencodegit clone --depth 1 https://github.com/rimio-ai/rimzWhat 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 | $0.00000 | $0.04352 |
| Opus 5 | $0.00000 | $0.02176 |
| Sonnet 5 | $0.00000 | $0.00870 |
| Haiku 4.5 | $0.00000 | $0.00435 |
Grade A, and why
adapter_opencode 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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenCode adapter
Read model.md for the provider-neutral agent model and adapter.md for the integration layer every adapter implements. Accounts, balances, and spend are in providers.md; the raw upstream protocol is in opencode-reference.md.
This doc is the single home for everything OpenCode-specific. OpenCode's integration surface is an in-process TypeScript plugin, so the adapter ships plugin.ts, embedded at compile time and installed as ~/.config/opencode/plugin/rimz.ts. The plugin subscribes to OpenCode hooks and bus events, flattens them into a RimZ-authored snake-case envelope, and spawns rimz hooks feed --source opencode with RIMZ_AGENT_PID set to the OpenCode process. OpenCode 1.18.23 publishes native permission and question prompts as permission.asked and question.asked; their *.replied/question.rejected outcomes clear waiting and record the native answer while OpenCode's UI remains responsible for collecting it. The published permission.ask compatibility hook remains wired for older releases and reads RimZ's stdout to set OpenCode's output.status.
Hooks and lifecycle
Native surface → internal mapping; the upstream hooks, bus payloads, SQLite schema, and auth shape are in opencode-reference.md.
| Native surface | Plugin emits | Channel | decode_hook → LifecycleSignal |
Normalized fields |
|---|---|---|---|---|
session.created without parentID |
session_created |
lifecycle | Registered |
worktree from directory/cwd |
chat.message |
chat_message |
lifecycle | TurnStarted |
sanitized prompt, model, effort |
session.idle root after a plan-agent turn |
session_idle + plan_proposed: true |
awaiting-user (PlanApproval) |
AwaitingInput { PlanApproval } → waiting |
latest token gauge; the native TUI owns review and mode switching |
session.idle root |
session_idle |
lifecycle | TurnEnded { errored: false, parked_on_background: false } |
latest token gauge stamped from message.updated |
session.error root |
session_error |
lifecycle | TurnEnded { errored: true, parked_on_background: false } |
error_class, error_message |
tool.execute.after root mutating tool |
tool_after |
lifecycle | ToolUsed { mutates: true, edits } |
edit/write/apply_patch/patch edit files; bash mutates only |
session.created with parentID; first child gauge update per model |
SubagentStart |
lifecycle | SubagentStarted |
child session_id, parent_session_id, sanitized title as task at creation; resolved model and tokens on gauge re-emit |
session.idle / session.error child |
SubagentStop |
lifecycle | SubagentStopped { errored } |
child session_id, parent_session_id |
experimental.session.compacting |
session_compacting |
lifecycle | Compacting |
leading compaction head |
session.compacted |
session_compacted |
lifecycle | CompactionEnded { auto: None } |
clears the compaction head |
session.deleted root |
session_ended |
lifecycle | Ended → stamps ended_at; runtime hides the retained row (ends_session) |
reason: "deleted" |
dispose root sweep |
session_ended |
lifecycle | Ended → stamps ended_at; runtime hides the retained row (ends_session) |
reason: "dispose"; all tracked roots, bounded to 1.5 seconds overall |
permission.asked bus event; legacy permission.ask hook |
permission_ask |
awaiting-user (Permission) |
AwaitingInput { Permission } → waiting |
permission name and matched patterns; the native prompt owns the answer |
question.asked |
question_ask |
awaiting-user (Question) |
AwaitingInput { Question } → waiting |
full questions, headers, options, multi-select/custom flags, joined title, and request id |
permission.replied |
permission_replied |
lifecycle | ToolUsed { mutates: false, edits: false } → clears waiting |
request id and once/always/reject answer recorded |
question.replied |
question_replied |
lifecycle | ToolUsed { mutates: false, edits: false } → clears waiting |
request id and per-question choices recorded |
question.rejected |
question_rejected |
lifecycle | ToolUsed { mutates: false, edits: false } → clears waiting |
request id and rejected outcome recorded |
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 · 91 lines · 0 tokens per session scan A 00c0a83e68f4
adapter_opencode is an agent published in the GitHub repository rimio-ai/rimz (23 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 4,352 tokens. 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 agents, from other repositories
maintaining-agent-docs
Agent "maintaining-agent-docs" from martin-francois/symphony-trello, covering maintaining agent docs, progressive disclosure contract, what may be added to agents.md (allowlist), what must not be added to agents.md (denylist) and where to put new guidance (decision tree).
progressive-disclosure-source
This file is the source of truth for the structure and maintenance approach.
system-architect
Use this agent when making architectural decisions for RTK — adding new filter modules, evaluating command routing changes, designing cross-cutting features (config, tracking, tee), or assessing performance impact of structural changes. Examples: designing a new filter family, evaluating TOML DSL extensions, planning…
documentation-analyst-writer
Use this agent when you need to analyze existing documentation and create new or updated documentation that strictly adheres to project-specific documentation standards defined in claude.md. This agent excels at maintaining consistency with established documentation patterns, ensuring technical accuracy, and producing…
qwen
@qwen-code/qwen-code is Alibaba's coding CLI built on top of Gemini CLI, tuned for the Qwen3-Coder family of models. adapters drives it via the qwen binary.
cursor
Adapter for the Cursor editor's agent CLI.