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 instructions/eugenepyvovarov/codeagentsmobile/agents-mdgit clone --depth 1 https://github.com/eugenepyvovarov/CodeAgentsMobileWrote 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/instructions/eugenepyvovarov/codeagentsmobile/agents-md)<a href="https://agentmods.dev/instructions/eugenepyvovarov/codeagentsmobile/agents-md"><img src="https://agentmods.dev/badge/instructions/eugenepyvovarov/codeagentsmobile/agents-md.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 | $0.03508 | $0.03508 |
| Opus 5 | $0.01754 | $0.01754 |
| Sonnet 5 | $0.00702 | $0.00702 |
| Haiku 4.5 | $0.00351 | $0.00351 |
Grade A, and why
CodeAgentsMobile AGENTS.md 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 3d 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 — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Guidelines
Runtime note (important)
- Claude Code chat is retired. OpenCode is the only chat runtime (
OpenCodeClient,OpenCode*Service,CodingAgentRuntimeKind.openCode). - Claude → OpenCode migration runs automatically per project via
ClaudeToOpenCodeMigrationServiceon chat open / send / runtime apply: promotes runtime, clears proxy transport anchors, copies compatible API keys, soft-copiesCLAUDE.md→AGENTS.mdwhen missing, imports.mcp.jsonintoopencode.json(c), ensures session + managed scheduler MCP, stampsopenCodeMigrationVersion. - Nil/unknown
agentRuntimeRawValueresolves to OpenCode (not Claude). - CodingAgentRuntimeRegistry always returns
OpenCodeRuntimeService(legacy.claudeProxyenum case remains for decode/migration only). - Agent daemon (
:8787,ProxyTaskService,server/claude-proxy) remains for tasks/push/scheduled runs until rehomed — do not delete those paths when cleaning chat. - Daemon Anthropic auth preference lives in
AgentDaemonAuthService(not Claude chat). - Do not expand Claude Code chat UX. MCP UI/management is OpenCode-primary (
CodingAgentMCPService→OpenCodeMCPService). - Chat must not import
ClaudeCodeService. Residual CCS usage is tasks/installer/legacy settings only until retirement.
ChatViewModel module map
| File | Responsibility |
|---|---|
ChatViewModel.swift |
Lifecycle, configure, public API, thin orchestration |
ChatViewModel+OpenCodeSend.swift |
Send / stream / abort |
ChatViewModel+OpenCodeHydration.swift |
Hydrate + session reconcile |
ChatViewModel+MCP.swift |
MCP cache / fetch / deferred |
ChatViewModel+ToolApproval.swift |
Tool permission + OpenCode questions |
ChatViewModel+Persistence.swift |
Load/save messages |
ChatViewModel+MediaPrefetch.swift |
CodeAgents UI media prefetch |
ChatViewModel+DeferredStartup.swift |
Post-ready MCP/rules/media queue |
ChatViewModelSupport.swift |
Pure helpers (merge, planners) |
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.
- 3d ago First seen · 166 lines · 3,508 tokens per session scan A 3933e727defa
CodeAgentsMobile AGENTS.md is an instructions file published in the GitHub repository eugenepyvovarov/CodeAgentsMobile (150 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 3,508 tokens to every session, about $0.0175 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 instructions, from other repositories
proxy CLAUDE.md
Claude Code instructions for routatic/proxy, covering claude.md, commands, start proxy with dashboard (recommended), start proxy only (headless) and architecture.
django-ai-plugins AGENTS.md
AGENTS.md instructions for vintasoftware/django-ai-plugins, covering contributor instructions, source of truth, changing a plugin, required validation and distribution policy.
PowerHou53 AGENTS.md
AGENTS.md instructions for Sdinzsh/PowerHou53, covering powerhouse 3 — shared agent rules (agents.md), tool-calling discipline, powerhouse 3 universal protocol, execution speed rules (highest priority) and think before act (non-trivial actions only).
plannotator AGENTS.md
AGENTS.md instructions for backnotprop/plannotator, covering plannotator, project structure, server runtimes, installation and environment variables.
solo AGENTS.md
Instructions for solo-agent/solo, covering project testing rules and project service lifecycle rules.
untether GEMINI.md
Instructions for littlebearapps/untether, covering untether — gemini instructions, stack & conventions, architecture, key rules and testing.