flutter-agentic AGENTS.md

Repository-specific instructions for FlutterAgentic, a Flutter application organized as a monorepo, meaning one repository containing multiple related parts. They point to architecture, feature, testing, and build documentation.

In plain words
What is it for?
Use them when adding features, creating use cases, designing screen state, reviewing code, changing app identifiers, or building the project.
Why use it?
They give coding agents the project structure and coding rules required before modifying the app.

Instructions file for CodexOpenCode

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/abhinav503/flutter-agentic/agents-md
Clone the repo
git clone --depth 1 https://github.com/abhinav503/flutter-agentic

Made for: Codex, OpenCode.

Per session 3,872 This file is loaded in full into every session.
When invoked 3,872 The same file — it is already loaded in full.
Security scan B 1 finding. Scan, not verified.
Origin 94% copy Near-identical to another mod 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.

ModelPer sessionOnce invoked
Fable 5 $0.03872 $0.03872
Opus 5 $0.01936 $0.01936
Sonnet 5 $0.00774 $0.00774
Haiku 4.5 $0.00387 $0.00387

Measured 3d ago against content hash be0da669da98, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

flutter-agentic AGENTS.md 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 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.

Enumerates other installed skillsmediumAgent snooping

Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.

**Skills live in two parallel trees — keep them in sync.** Both Claude Code (`.claude/skills/`) and Codex CLI (`.codex/skills/`) read per-skill `SKILL.md` folders, but in different formats: Claude skills are thin (body +
Origin

This is a copy

94% identical to flutter-agentic copilot-instructions.md — 58 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

AGENTS.md · 186 lines

How it starts

The opening of the file, as written. The whole thing — 186 lines — stays where its author put it; the contents beside it link to each section on GitHub.

FlutterAgentic — Agent Rules (Android Studio · Codex CLI)

Full reference → docs/ folder · Setup → $setup-project · Release → $release

Documentation Index

Read before writing or modifying any code:

  • docs/reference/architecture.md — core folder map, layer patterns, naming, DI, error flow, design system, testing
  • docs/explanation/end-goal.md — project vision and guiding principles

Read on demand:

  • docs/how-to/contributing.md — contributor workflow and git hooks
  • docs/how-to/add-feature-template.md — full folder tree, empty class skeletons, DI wiring, and forbidden-pattern checklist for scaffolding a new feature
  • docs/how-to/add-usecase.md — create a use case class and register it in injection_container.dart
  • docs/how-to/design-screen-state.md — business-logic naming for events and states, retry context rules, screen rendering pattern; use the jokes feature as the reference
  • docs/how-to/review-code.md — when asked to review, audit, or check generated code; run through the full checklist and report ✅/❌ per section
  • docs/how-to/change-app-id.md — when asked to change the application ID or bundle identifier; covers Android (build.gradle.kts + MainActivity.kt package path) and iOS (project.pbxproj), with Xcode manual steps and provisioning notes
  • docs/how-to/rename-app.md — when asked to rename the app; covers display name, package name, and all files that reference the old name
  • docs/how-to/connect-firebase.md (or run $connect-firebase) — when connecting an app to Firebase; covers checking/installing the Firebase + FlutterFire CLIs, running flutterfire configure, per-app firebase_core, main.dart init, Android Gradle plugin, iOS deployment target (15.0+), and the Xcode GoogleService-Info.plist registration check
  • docs/explanation/ai-agents.md — per-agent install and usage
  • docs/tutorials/solid-principles.md — how SOLID principles are applied across all layers; useful when designing new classes or reviewing layer boundaries
  • docs/tutorials/design-patterns-and-concepts.md — design patterns used in this codebase (Singleton, Repository, DTO, Either, Sealed Classes, Strategy, and more)
  • Release workflow ($release) — when asked to do a release, follow these steps interactively; ask for confirmation at each step before proceeding:
    1. Load GH_TOKEN from the git-ignored root .env (set -a && . ./.env && set +a) — release auth is explicit because the repo uses multiple GitHub accounts — then check gh auth status reports (GH_TOKEN). Source .env in every shell that runs gh. If .env lacks a token, create a fine-grained PAT (Contents: Read and write) at https://github.com/settings/personal-access-tokens/new and add GH_TOKEN=…. Stop if not ready. If gh auth status reports the token as invalid, confirm the shell has network access before replacing it — in a sandboxed agent environment, blocked network access can surface as an auth failure; re-run with network permission and .env sourced first.
    2. Get current branch (git branch --show-current). Confirm release branch with user.
    3. Compare to main: git log main..{BRANCH} --oneline + git diff main..{BRANCH} --stat. Show commits.
    4. Read version (grep "^version:" pubspec.yaml). Propose bump: Major = breaking; Minor = feat: or new component/skill; Patch = fix/chore/docs. Wait for confirmation.
    5. Edit pubspec.yaml with confirmed version.
    6. Create docs/releases/v{VERSION}.md from docs/releases/_template.md. Two sections: Features (what developers gain) and Agent Context Improvements (what agents gain). Plain language, one sentence per bullet, no duplicates. Show draft and wait for confirmation.
    7. Commit: git add pubspec.yaml docs/releases/v{VERSION}.md && git commit -m "chore: release v{VERSION}" && git push
    8. Merge: git checkout main && git pull origin main && git merge --no-ff {BRANCH} -m "chore: merge {BRANCH} into main for v{VERSION}" && git push origin main
    9. Tag and release: git tag v{VERSION} && git push origin v{VERSION} then gh release create v{VERSION} --title "v{VERSION} — {TITLE}" --notes-file docs/releases/v{VERSION}.md --target main. Report URL.
    10. Ask to delete release branch. If yes: git branch -d {BRANCH} && git push origin --delete {BRANCH}

Read the full file on GitHub · 186 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.

  1. 3d ago First seen · 186 lines · 3,872 tokens per session scan B be0da669da98

Subscribe to this mod's changes

flutter-agentic AGENTS.md is an instructions file published in the GitHub repository abhinav503/flutter-agentic (10 stars, last pushed 10d ago), licensed MIT. It adds 3,872 tokens to every session, about $0.0194 per session on Opus 5. A static security scan graded it B with 1 finding (enumerates other installed skills). It is 94% identical to flutter-agentic copilot-instructions.md, differing in 58 lines, and is treated as a copy.