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/randomittin/heimdallWrote 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/commands/randomittin/heimdall/maintain)<a href="https://agentmods.dev/commands/randomittin/heimdall/maintain"><img src="https://agentmods.dev/badge/commands/randomittin/heimdall/maintain/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/commands/randomittin/heimdall/maintain"><img src="https://agentmods.dev/badge/commands/randomittin/heimdall/maintain.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 | $0.00040 | $0.01419 |
| Opus 5.5 | $0.00016 | $0.00568 |
| Sonnet 5.5 | $0.00008 | $0.00284 |
| Haiku 4.5 | $0.00004 | $0.00142 |
Grade A, and why
maintain 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 1mo 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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Maintain — Automated Bug Detection & Fix Pipeline
Two-phase maintenance: seek then fix.
Usage
/hmd:maintain — run once (seek + fix)
/hmd:maintain seek — only find bugs, raise issues
/hmd:maintain fix — only fix existing issues
/hmd:maintain auto — schedule recurring via /schedule
This pipeline vs the engine-driven autopilot
This file describes the prose seek/fix pipeline: two agents (seeker, fixer)
making direct gh CLI calls. It does not use bin/heimdall-issue-queue or
bin/lib/connectors/ — the seeker files issues straight to GitHub with gh issue create, and the fixer lists them straight back with gh issue list --label bug --state open. The shared bug label is the only "queue": whatever carries it is
in scope for the next fixer pass, oldest first.
A separate, engine-driven autopilot (/hmd:maintain-check, backed by
bin/heimdall-maintain-loop → bin/heimdall-issue-loop) uses the real
normalized, prioritized, multi-source queue (bin/heimdall-issue-queue, piece b
of the issue-resolution-loop design) — but its own GitHub sync
(sync_queue_from_github in bin/lib/maintain_loop.py) talks to gh directly
too, filtered on label maintainer. The seeker now applies that label
(bug,seeker,maintainer — see agents/seeker.md), so a seeker-filed issue IS
ingested by the engine. Until 2026-08-23 it was not: seeker filed bug,seeker
and the engine requested maintainer, and gh's --label filter is a
superset/AND match rather than an OR, so the two sets never intersected and the
seek-then-fix loop had never once run end to end.
The two pipelines remain independent in every other respect — this prose one
discovers via its own gh issue list, and running it does not drive the engine.
Connectors — pluggable, but not in THIS pipeline
bin/lib/connectors/ (github/slack/email/corpus) + bin/heimdall-issue-queue
are real and working — proven by direct execution, not just by reading the
source — but they back the engine autopilot above, not the prose pipeline in
this file. To feed a non-GitHub source into that engine's queue by hand:
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.
- 1mo ago Changed · +40 lines ffad7d1b12b3
- 1mo ago First seen · 81 lines · 40 tokens per session scan A 2d3cd6f71e63
maintain is a command published in the GitHub repository randomittin/heimdall (5 stars, last pushed 4d ago), licensed MIT. It adds 40 tokens to every session and 1,419 once invoked, about $0.0002 per session on Opus 5.5. 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 commands, from other repositories
subagent-diagnose
Multi-agent failure-investigation orchestrator. Parallel to /subagent-implementation — same scratchpad + investigator + builder/surgeon + reviewer + FOLLOWUPS pattern — but starts from a failure, not a spec. Two modes share one loop: ci (failed CI run is the brief seed) and bug (freeform symptom paragraph is the brief…
quick-fix
Implement→review subagent loop with no planning phase, for a fix whose cause is known and whose shape is obvious, however many files it touches. No spec, no worktree, no docs or signals at the end; one audit. Routes to /subagent-diagnose on an unknown cause and to /atomic-plan on an open approach or contract choice.
atomic-help
Route a lost user to the right atomic verb, skill, agent, or binary subcommand. Bare invocation reads git state and recommends one next step. Topic keyword or freeform intent → focused pointer. tour → guided walkthrough of the system. Help router, not duplicated docs.
follow-up
Review and act on pending reminders. Bare invocation shows all reminders as an indexed list; cron-fired invocation (/follow-up due ) surfaces the specific reminder and waits for a response; /follow-up review lists stale project follow-up entries for per-item disposition (extend/close/promote/skip). Transport-aware…
review-branch
Review the current branch's diff against base by dispatching atomic-reviewer. No orchestration loop, no spec required — pre-flight before /commit pr or /commit merge.
report-issue
Open a atomic GitHub issue via gh. Bug report or feature request — auto-detect from user's description.