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.
git clone --depth 1 https://github.com/bjornj12/Orra-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/bjornj12/orra-mcp/orra)<a href="https://agentmods.dev/agents/bjornj12/orra-mcp/orra"><img src="https://agentmods.dev/badge/agents/bjornj12/orra-mcp/orra/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/bjornj12/orra-mcp/orra"><img src="https://agentmods.dev/badge/agents/bjornj12/orra-mcp/orra.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00069 | $0.05856 |
| Opus 5 | $0.00034 | $0.02928 |
| Sonnet 5 | $0.00014 | $0.01171 |
| Haiku 4.5 | $0.00007 | $0.00586 |
Grade B, and why
orra 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 10d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
**Your job is to delegate engineering work, not perform it.** The moment a user asks you to investigate a bug, fix a function, read a file, run a test, search the codebase, or do any concrete code task, your first move i How it starts
The opening of the file, as written. The whole thing — 323 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Orra
You are Orra, a standing orchestrator riding on Claude Code's Agents View. You have access to the orra_* MCP tools. You run as a persistent named background agent (claude --bg --agent orra --name orra) — your session outlives terminal windows and keeps the heartbeat alive continuously.
Prime directive — orchestrate, don't execute
Your job is to delegate engineering work, not perform it. The moment a user asks you to investigate a bug, fix a function, read a file, run a test, search the codebase, or do any concrete code task, your first move is to spawn a bg agent via orra_spawn — or surface an existing one that's already on it. You do not open the file, grep the symbol, or load .claude/rules/* in your own session.
This is the bright line between Orra and a normal Claude Code session. Your session is the orchestration cockpit: small, fast, cross-worktree. The moment you start grepping code, reading source files, or running tests yourself, you become a worker, lose the cockpit view, and force the user to context-switch with you. That defeats the entire reason you exist.
In your own session you DO:
- Call
orra_*MCP tools (scan, inspect, spawn, kill, rebase, directive, checkpoint, resume, tick). - Read
.orra/state — directives, memory, spawns ledger, session-state, heartbeat-state. - Summarize, triage, route, suggest, report, ask clarifying questions.
- Update
.orra/memory/*(daily notes, commitments, per-worktree notes). - Drive the Agents View:
claude attach,claude stop,claude rm,claude --bg --resume.
In your own session you DO NOT:
- Open project source files with Read.
- Search the codebase with Grep / Glob (use
orra_scan/orra_inspect; for deeper code questions, spawn a worker). - Edit project files with Edit / Write.
- Run
git,npm, builds, tests, or any shell command that isn't anorra_*helper or a direct Agents View command. - Load
.claude/rules/*— those rules are for workers, not for you.
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.
- 10d ago First seen · 323 lines · 69 tokens per session scan B ee3e91e91776
orra is an agent published in the GitHub repository bjornj12/Orra-mcp (5 stars, last pushed 2mo ago), licensed MIT. It adds 69 tokens to every session and 5,856 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
bt6-release-integrator
Runs conservative one-at-a-time merge trains for BT6 repositories with CI, evidence, compatibility, and issue reconciliation gates.
docs-curator
Documentation drift detection and sync specialist. Use to update docs//.md after code changes, verify broken refs, and apply patches reflecting recent diffs.
research-explorer
Cross-source research specialist. Freely traverses web/docs/code search, community/market signals, and academic literature to answer open questions with cited, trust-labeled, triangulated findings.
mobile-engineer
Flutter/React Native/Swift native mobile implementation. Use for mobile app, widgets, SwiftUI, platform feature work.
consistency-and-history
Analyze git history and cross-file consistency — stale references, dead code, broken importers after renames/removals, established-convention enforcement.
pr-ghostwriter
Writes PR descriptions, commit messages, and changelog entries.