What is codeagent-mobile-clients AGENTS.md?

Repository instructions for the client-side code of Codeam Mobile Clients. The project contains a command-line tool, editor extensions, a JetBrains plugin, and shared communication code.

Use them when working on the repository, especially client releases, shared protocol changes, real-time output handling, and type-safe TypeScript edits.

They give coding agents the rules needed to change this multi-package project safely, including release, shared-code, real-time event, and TypeScript requirements.

codeagent-mobile-clients AGENTS.md — Instructions file for Codex and OpenCode by edgar-durand: Repository instructions for the client-side c

Install

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.

agentmods

npx agentmods add instructions/edgar-durand/codeagent-mobile-clients/agents-md

Clone the repo

git clone --depth 1 https://github.com/edgar-durand/codeagent-mobile-clients

Made for: Codex, OpenCode.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/edgar-durand/codeagent-mobile-clients/agents-md.svg)](https://agentmods.dev/instructions/edgar-durand/codeagent-mobile-clients/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/edgar-durand/codeagent-mobile-clients/agents-md"><img src="https://agentmods.dev/badge/instructions/edgar-durand/codeagent-mobile-clients/agents-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 846 This file is loaded in full into every session.

When invoked 846 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00846	$0.00846
Opus 5	$0.00423	$0.00423
Sonnet 5	$0.00169	$0.00169
Haiku 4.5	$0.00085	$0.00085

Measured 4d ago against content hash bc4d78d73c3a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codeagent-mobile-clients 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 4d 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.

AGENTS.md · 30 lines

How it starts

The opening of the file, as written. The whole thing — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Canonical agent instructions for this repo live in CLAUDE.md. Read it first — this file is a cross-tool entry point that points there and surfaces the highest-signal rules inline.

This is the client-side repo: codeam-cli (npm), the VS Code / Cursor / Windsurf extension, the JetBrains plugin, and @codeam/shared (the wire protocol). The backend lives in the sibling codeagent-mobile repo.

Non-negotiables (full detail in CLAUDE.md)

Releases are tag-triggered: all three clients ship under ONE version line. git tag vX.Y.Z && git push origin vX.Y.Z publishes npm + VS Code Marketplace + Open VSX + JetBrains. Commits to main run CI but do NOT release.
Protocol code is shared: parsing / pricing / chunk shapes change ONLY in packages/shared (@codeam/shared); both CLI and VS Code import through it.
No polling for realtime: react inside the PTY/SSE event (OutputService.push), don't add setTimeout "check again" loops. The HTTP-polling fallback in command-relay.service.ts is the one documented exception.
Typing: no as unknown as / any to silence TS — fix the source.

Agent-failure messaging (2026-06-24 — `apps/cli/src/agents/acp/runner.ts`)

A turn must NEVER end silently or with a misleading status. Tests: __tests__/agents/acp.failureBubble.test.ts.

failureBubble is the sole arbiter, keyed on the agent's OWN error.
The provider-outage bubble fires ONLY from the agent's error — NEVER from polling the status page. A status-page incident can be live while the user's local agent is fine. The checkProviderStatus status-page catch-all was removed in v2.42.0 for that false positive. The status page is the link inside the bubble, not a trigger.
Auth notice as a completed-turn reply (Not logged in · Please run /login printed as plain text, turn ends cleanly) → caught by replyIsAuthFailure (≤200-char guard) → re-auth bubble + reportCredentialInvalid.
1M-context usage-credits gate (v2.43.0): claude Code v2.1.x always sends the context-1m beta even on credit-less accounts → 429 "Usage credits required for 1M context" every turn (NOT Headroom — it forwards the beta unchanged). On-demand recovery only: looksLike1mContextCreditsError → offer a "Disable 1M context and continue" select_prompt → on tap, persist disable1mContext, re-spawn the ACP adapter with CLAUDE_CODE_DISABLE_1M_CONTEXT=1, re-run the failed prompt. Never disabled for all users.

Read the full file on GitHub · 30 lines

Changes

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.

4d ago First seen · 30 lines · 846 tokens per session scan A bc4d78d73c3a

Subscribe to this mod's changes

codeagent-mobile-clients AGENTS.md is an instructions file published in the GitHub repository edgar-durand/codeagent-mobile-clients (9 stars, last pushed 4d ago), licensed MIT. It adds 846 tokens to every session, about $0.0042 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.

ai claude claude-code cli copilot cursor

codeagent-mobile-clients AGENTS.md

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

AGENTS.md

Non-negotiables (full detail in CLAUDE.md)

Agent-failure messaging (2026-06-24 — `apps/cli/src/agents/acp/runner.ts`)

What this file has done since we first saw it

Other instructions, from other repositories

mcp-unity AGENTS.md

rails-ai-context AGENTS.md

inspecto CLAUDE.md

skill.color-expert CLAUDE.md

imagine-mcp AGENTS.md

llm-ide-rules typescript.instructions.md

codeagent-mobile-clients AGENTS.md

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

AGENTS.md

Non-negotiables (full detail in CLAUDE.md)

Agent-failure messaging (2026-06-24 — apps/cli/src/agents/acp/runner.ts)

What this file has done since we first saw it

Other instructions, from other repositories

mcp-unity AGENTS.md

rails-ai-context AGENTS.md

inspecto CLAUDE.md

skill.color-expert CLAUDE.md

imagine-mcp AGENTS.md

llm-ide-rules typescript.instructions.md

Agent-failure messaging (2026-06-24 — `apps/cli/src/agents/acp/runner.ts`)