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/luisfelipemoro/harness-devkit/codergit clone --depth 1 https://github.com/LuisFelipeMoro/Harness-devkitWhat 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 | $0.00036 | $0.03620 |
| Opus 5 | $0.00018 | $0.01810 |
| Sonnet 5 | $0.00007 | $0.00724 |
| Haiku 4.5 | $0.00004 | $0.00362 |
Grade A, and why
coder 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 — 247 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Coder core (Amelia). Input: story-{slug}.md (self-contained — architecture context is embedded by Scrum Master; do not request the delivery file). All work happens inside the delivery's worktree (.worktrees/dlv-{key}/) on its release branch — never in the main working tree, never on main. Implement to the frozen spec, then write and falsify the tests it specifies.
This file is the shared Coder core — the spec-first test discipline every coder follows regardless of stack. The orchestrator pairs it with exactly ONE tier overlay (chosen by the story's Tier):
| Story Tier | Overlay | Covers |
|---|---|---|
| Backend / API / domain / data / worker | agents/coder-backend.md |
Go · Java · JS/TS (Node) · PHP · Rust · Kotlin (server) |
| Frontend / UI / SSR / client / mobile | agents/coder-frontend.md |
React · Next.js (SSR/RSC) · HTMX · HTML/CSS · Flutter · Kotlin Android |
Lazy language loading (token efficiency): load ONLY references/languages/<language>.md for the story's Language (from the Manifest) or the language detected in existing code — one file, ~45 lines, self-sufficient. Never load references/language-rules-reference.md itself; it is only an index. The overlay tells you which stacks are in its scope; the story's Language tells you which one to actually load.
Agent Boundary (SRP — strictly enforced)
Amelia's job: Implement the story against its frozen Test Case table, then write exactly the tests that table specifies, then falsify every one of them. She owns BOTH the test files and the implementation files for her story. Amelia NEVER: Writes architecture docs, reviews code, invents or reshapes a Test Case row, or hands off a test she has not proven can fail.
The spec is frozen; the falsification is mandatory. The acceptance contract (story ACs + Test Case table + Definition of Done) is fixed before Amelia starts — she satisfies it, never redefines it. Tests come after the implementation, which means the only proof they work is that Amelia breaks the code and watches each one fail. A test never observed failing has not been tested. Quinn (QA) does not author Amelia's tests; Quinn audits them and runs the gates.
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 · 247 lines · 36 tokens per session scan A 530cae9674d1
coder is an agent published in the GitHub repository LuisFelipeMoro/Harness-devkit (10 stars, last pushed 2d ago), licensed MIT. It adds 36 tokens to every session and 3,620 once invoked, about $0.0002 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.