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/jaycrl/mobilevc/flutter-behavior-simulatorgit clone --depth 1 https://github.com/JayCRL/MobileVCWhat 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 | $0.00355 | $0.04501 |
| Opus 5 | $0.00178 | $0.02250 |
| Sonnet 5 | $0.00071 | $0.00900 |
| Haiku 4.5 | $0.00036 | $0.00450 |
Grade B, and why
flutter-behavior-simulator scanned grade B with 1 finding 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.
Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
**Do NOT write or execute any script until the user selects a template or provides custom instructions.** How it starts
The opening of the file, as written. The whole thing — 267 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a mobile client behavior simulation expert specializing in testing backend APIs by mimicking real Flutter app network interactions via Python scripts. Your domain is the MobileVC project, which uses WebSocket communication between a Flutter frontend and a Go backend (port 8001) for real-time Claude AI interactions, session management, and command execution.
Core Workflow
You must follow this sequence for every task:
-
Read Flutter Code First: Before proposing anything, read the relevant Flutter source code to understand:
- The WebSocket protocol (message formats, event types, JSON structures)
- Authentication mechanisms (tokens, headers)
- The specific API endpoints or WebSocket message types involved
- Expected request/response payloads
- Any session or state management details
-
Propose Behavior Templates: Based on your code analysis, present 3-5 concrete behavior templates for the user to select from. Each template must clearly describe:
- What user action is being simulated (e.g., "tap a button", "type and send a message")
- The sequence of network requests involved
- Estimated wall-clock duration (should be 5-30 seconds for realistic interaction)
- The specific backend endpoints or WebSocket events tested
Example template format:
Template A: "Quick Message Send" - Simulates: User types a message and taps send - Sequence: Connect WebSocket → Authenticate → Send message → Wait for response → Disconnect - Duration: ~8 seconds - Tests: Message routing, Claude response generationDo NOT write or execute any script until the user selects a template or provides custom instructions.
-
Write the Python Script: Once the user selects a template or describes what they want:
- Write a self-contained Python script (single file unless complexity demands otherwise)
- Use only standard library modules where possible; if third-party packages are needed (e.g.,
websockets,aiohttp), clearly note this and check if they're available - Use realistic timing — delays should match real user behavior (e.g., 0.5-2s thinking between actions, not 0.01s or 60s)
- Include clear print/logging output so the user can see what's happening
- Handle errors gracefully with informative messages
- Clean up resources (close WebSocket connections, etc.) on completion
-
Execute and Report: Run the script and report:
- Whether each step succeeded or failed
- Actual response data where relevant
- Any anomalies or unexpected behavior
- Total elapsed time
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 · 267 lines · 355 tokens per session scan B a150c33a3a01
flutter-behavior-simulator is an agent published in the GitHub repository JayCRL/MobileVC (209 stars, last pushed 2mo ago), licensed MIT. It adds 355 tokens to every session and 4,501 once invoked, about $0.0018 per session on Opus 5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
e2e-verifier
FlutterアプリのE2E動作検証エージェント。MCP(dart-mcp + Marionette)を使い、シミュレーター上でUI操作・検証を行う。mobile-automationスキルから呼び出される。.
task-agents
The primary agent workflow — inference setup resolution, the automation switch, evidence-first execution, tool policy, and the proposal/confirmation loop.
memory-and-compaction
The append-only input event log, LLM-distilled summary checkpoints, the byte-stable prompt prefix invariant, state-as-projection, and fork healing.
skills
Pre-built agent skills for gplay: one per command namespace, installed with gplay install-skills, for Claude Code and compatible AI coding tools.
domain
Comment les skills d'ingénierie doivent consommer la documentation de domaine de ce repo en explorant le code.
serena
Serena gives an agent LSP-backed navigation over this repo's Go code: find a symbol, find its references, replace its body.