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 skills add randomittin/heimdall --skill system-healthgit 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/skills/randomittin/heimdall/system-health)<a href="https://agentmods.dev/skills/randomittin/heimdall/system-health"><img src="https://agentmods.dev/badge/skills/randomittin/heimdall/system-health/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/skills/randomittin/heimdall/system-health"><img src="https://agentmods.dev/badge/skills/randomittin/heimdall/system-health.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 | $0.00102 | $0.01884 |
| Opus 5 | $0.00051 | $0.00942 |
| Sonnet 5 | $0.00020 | $0.00377 |
| Haiku 4.5 | $0.00010 | $0.00188 |
Grade A, and why
system-health 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 8d 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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
system-health
A read-only advisor: does this machine need a cleanup, and is Heimdall causing it? It grades disk + memory pressure, counts Heimdall's own leaked python, and SUGGESTS the fix. It never deletes or kills anything unless you explicitly ask for the scoped self-reap.
Pairs with mac-deep-clean (the executor that actually reclaims disk) — a separate,
user-installed skill, NOT bundled with Heimdall. This skill is the trigger — the thing
that notices you need it — and it checks whether mac-deep-clean is actually installed
(~/.claude/skills/mac-deep-clean/SKILL.md or the project-level equivalent) before ever
naming it, falling back to plain read-only investigation commands when it's absent.
Why this exists (the lesson it encodes)
A long Heimdall session leaks background python: the presence keeper's presence-doctor
spawns mock_cp.py (a mock control-plane) children that get orphaned to launchd (ppid 1)
and are never reaped. One session accumulated ~692 of them, pinning all 16G of RAM and
thrashing swap 17G/17G. heimdall-gc reaps orphaned keeper pidfiles, but NOT these
launchd-reparented children (nothing tracks them) — so they slip the collector. Most of a
real cleanup session was just purging these do-nothing python procs. This advisor is the
check that catches them early.
Run it — the unified core (both axes at once)
bin/heimdall-cleanup is the ONE entrypoint that covers BOTH axes — MEMORY (RAM/swap +
hmd's orphan leak) and DISK (hmd's own reclaimable garbage). It ORCHESTRATES the two
reapers below; it never re-implements them. Reach for it first:
heimdall-cleanup # report (DEFAULT, read-only): unified health + reclaimable-now MB
heimdall-cleanup --json # {severity, memory, disk, procs, reclaim} for programmatic use
heimdall-cleanup --apply # reap hmd's OWN garbage only (orphaned hmd python + gc disk)
heimdall-cleanup --deep # SUGGEST-only handoff to mac-deep-clean (confirm-gated, nothing auto)
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.
- 8d ago First seen · 124 lines · 102 tokens per session scan A de28090db066
system-health is a skill published in the GitHub repository randomittin/heimdall (5 stars, last pushed 2d ago), licensed MIT. It adds 102 tokens to every session and 1,884 once invoked, about $0.0005 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 skills, from other repositories
atomic-debug
Hypothesis-driven debugging skill. Use when a bug, test failure, crash, or unexpected behavior is reported. Auto-trigger on error pastes or "broken/doesn't work/failing" language. Explicit invocation: /atomic-debug. Output: symptom statement → hypothesis table → cheapest test first → root cause. No symptom-patching.…
memstack-development-performance-audit
Use this skill when the user says 'performance audit', 'why is it slow', 'optimize performance', 'page speed', 'Core Web Vitals', 'lighthouse', 'load time', or needs to diagnose and fix frontend or backend performance issues. Do NOT use for code reviews or security audits.
atomic-writing
One voice for every file the repo ships: README, docs/guides/, docs/reference/, docs/spec/, docs/design/, docs/research/, docs/wiki/, CLAUDE.md, and the prompt artifacts under commands/, agents/, skills/, rules/, and output-styles/. Clear, direct, technical, and visual wherever the content has a shape. No marketing…
atomic-documentation
Diff-driven documentation surface classifier. Given a diff (staged, branch, or range), reads the project's indexed ## Documentation surfaces table from CLAUDE instructions, matches the diff against it, and emits a structured list of proposed edits. Two modes: maintenance (commit flow — stale/incomplete only, never…
atomic-visual-options
Planning-phase visual comparison aid. Renders 2-4 side-by-side variants per decision dimension as a single throwaway, self-contained HTML file and captures the user's pick as typed terminal codes (e.g. "A2 B3"). Auto-fires on phrases like "show me a few options", "mock up some variants", "let me see this side by…
atomic-git-discipline
Compressed commit message and PR body generator. Cuts noise from both while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious; PR bodies state only what the diff can't show, 120 words max. Use when user says "write a commit", "commit message", "generate…