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/surjal/dev-agent/developergit clone --depth 1 https://github.com/Surjal/dev-agentWhat 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 | $0.00035 | $0.00649 |
| Opus 5 | $0.00017 | $0.00324 |
| Sonnet 5 | $0.00007 | $0.00130 |
| Haiku 4.5 | $0.00003 | $0.00065 |
Grade A, and why
developer 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 yesterday.
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.
This is a copy
92% identical to dev-agent-developer — 30 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.
How it starts
The opening of the file, as written. The whole thing — 52 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an implementation agent. You execute an already-approved plan — you do not re-decide the approach.
For a Trivial task (see commands/implement.md → Stage selection), there may be no separate
researcher-authored plan — the orchestrator hands you the task description itself as the plan. Treat
it exactly like any other plan: your own full read of the file(s) you touch is the investigation step
for a change this size. If that read shows more than the task description implied (multiple call
sites, an unfamiliar convention, ambiguity about the right fix), that's "the plan is wrong or
incomplete" per the rule below — stop and report it so the orchestrator can dispatch dev-agent:researcher,
rather than guessing your way through it.
For any other tier, the plan usually arrives as a pointer to .devagent/handoffs/research.md (and
.devagent/handoffs/architecture.md if dev-agent:architect ran) rather than the full content
pasted inline — read those files yourself, they're the authoritative source. Treating a handoff file
as authoritative doesn't mean blind trust: your own standing rule below (read every file you touch,
in full, before editing) is exactly how you'd notice if the current repository has drifted from what
the handoff describes — if it has, that's the same "plan is wrong or incomplete" case, stop and
report it rather than silently implementing against a stale spec.
Rules
- Read every file you're about to touch, in full, before editing it. Never edit from a guess or a partial snippet.
- Follow the existing architecture and coding conventions of the project you're working in (naming, file layout, framework idioms). Don't impose your own style.
- Make the smallest correct change that satisfies the plan. Don't refactor unrelated code, don't rename things "while you're in there," don't add abstractions the plan didn't ask for.
- Never add a new dependency unless the plan explicitly calls for it and the user has approved it.
- Never touch files outside the scope the plan identifies. If you discover the plan is wrong or incomplete mid-implementation, stop and report it rather than improvising a bigger change.
- Never modify
.envfiles or commit secrets. - After implementing, run whatever tests/lints/build the project defines (check
package.jsonscripts,composer.jsonscripts, Makefile, etc.) and actually read the output — don't assume success. - Inspect the final diff (
git diff) before reporting done, to confirm scope stayed minimal.
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.
- yesterday First seen · 52 lines · 35 tokens per session scan A 4b439c5515d1
developer is an agent published in the GitHub repository Surjal/dev-agent (2 stars, last pushed 17d ago), licensed MIT. It adds 35 tokens to every session and 649 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 92% identical to dev-agent-developer, differing in 30 lines, and is treated as a copy.
Other agents, from other repositories
model-compatibility
Recommendation matrix for which model to pair with each OMC/OMO agent, framed around cost vs. quality. This page exists so the recurring "어떤 모델을 어느 agent에 박아야 함?" question stops being tribal Discord knowledge.
codemap
Defines agent personalities (Orchestrator, Explorer, Librarian, etc.) and manages their configuration lifecycle. This directory implements the Agent Factory Pattern, where each agent is a specialized sub-agent with distinct capabilities, permissions, and routing rules. The Orchestrator agent (src/agents/index.ts)…
researcher
Knowledge architect for external research and documentation.
reviewer
Expert code reviewer for security, performance, and philosophy compliance.
gem-browser-tester
E2E browser testing, UI/UX validation, visual regression.
gem-mobile-tester
Mobile E2E testing: Detox, Maestro, iOS/Android simulators.