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/innocent-children/dev-flow/agents-mdgit clone --depth 1 https://github.com/Innocent-children/dev-flowWhat 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 | $0.01856 | $0.01856 |
| Opus 5 | $0.00928 | $0.00928 |
| Sonnet 5 | $0.00371 | $0.00371 |
| Haiku 4.5 | $0.00186 | $0.00186 |
Grade A, and why
dev-flow 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 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 — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dev Flow Repository Instructions
Authority
Before implementation work, read in this order:
- the user's current explicit request and acceptance criteria;
CONTRIBUTING.mdorCONTRIBUTING_en.md;docs/PRODUCT.mdordocs/PRODUCT_en.md;- the technical documents directly related to the change;
- the current source code, schemas, package manifests, and executable tests for the affected surface.
Before a version-only release, read in this order:
release/README.md;- the selected product's release README under
release/; - the current prepare and publisher contracts under
release/; - the package manifest and current public-version metadata.
When documentation and executable behavior disagree, use the executable implementation to determine current behavior and update the affected documentation in the same change. Do not infer requirements from branch names, directory names, chat history, or historical design documents.
Requirement Scope
The user's current explicit instruction, current public contracts, and existing product boundaries define authorized product work.
- Every implementation task must map to the current request, an acceptance criterion, a public contract, or an approved engineering constraint.
- Identify the exact files or directories affected before implementation.
- Do not convert rationale, examples, future candidates, or historical incidents into new behavior.
- Do not broaden an implementation because a nearby abstraction appears useful.
- When the request conflicts with current contracts or leaves a material product choice unresolved, stop and ask for direction before changing behavior.
- Historical design material is available through Git history; it is not current implementation authority.
Documentation and Internationalization
Human-readable documentation mirrors delivered product behavior; it is not runtime, build, release,
or test authority. The maintained locale set and document-family coverage are defined by
docs/I18N.md and docs/I18N_en.md.
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 · 190 lines · 1,856 tokens per session scan A b99c72b27a1c
dev-flow AGENTS.md is an instructions file published in the GitHub repository Innocent-children/dev-flow (4 stars, last pushed 3d ago), licensed Apache-2.0. It adds 1,856 tokens to every session, about $0.0093 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.
Other instructions, from other repositories
codex-provider-sync AGENTS.md
Instructions for Dailin521/codex-provider-sync, covering ai / agent operator guide, vnext architecture baseline, goal, choose the interface and safe operating flow.
supergraph AGENTS.md
Instructions for datit309/supergraph, covering supergraph — mandatory workflows, skills, auto language detection, tiered workflow — pick the right tier first and full pipeline (tier 3).
frappe-agent copilot-instructions.md
Instructions for Dkm0315/frappe-agent: Treat this repository as Frappe, ERPNext, and Frappe ecosystem guidance for coding agents.
GameStudio-CodexKIT AGENTS.md
Instructions for hoatv2211/GameStudio-CodexKIT, covering gamestudio-codexkit operating contract, active sources of truth, primary distribution, repository-local maintenance and evidence.
codex-third-party-subagents AGENTS.md
Instructions for dhy365-creator/codex-third-party-subagents, covering agents.md, project scope, rules for agents working here, layout and validation.
codexhub AGENTS.md
Instructions for zpf2234/codexhub, covering codexhub contributor guide and project expectations.