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/jianzhichun/permafrostWrote 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/jianzhichun/permafrost/doctor)<a href="https://agentmods.dev/commands/jianzhichun/permafrost/doctor"><img src="https://agentmods.dev/badge/commands/jianzhichun/permafrost/doctor/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/jianzhichun/permafrost/doctor"><img src="https://agentmods.dev/badge/commands/jianzhichun/permafrost/doctor.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 | $0.00013 | $0.00328 |
| Opus 5 | $0.00006 | $0.00164 |
| Sonnet 5 | $0.00003 | $0.00066 |
| Haiku 4.5 | $0.00001 | $0.00033 |
Grade A, and why
doctor scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
allowed-tools: Bash(curl:*) What it actually says
Diagnose the DeepSeek prefix cache for this session.
Pull the proxy's alignment report and prefix-change history:
!curl -fsS "http://127.0.0.1:${PERMAFROST_PORT:-8787}/permafrost/doctor" 2>/dev/null || echo '{"error":"proxy not reachable — start it with: permafrost up"}'
Also check the two environment knobs that most often break the cache under a custom endpoint:
!echo "ANTHROPIC_BASE_URL=${ANTHROPIC_BASE_URL:-<unset>} ENABLE_TOOL_SEARCH=${ENABLE_TOOL_SEARCH:-<unset>} PERMAFROST_MODE=${PERMAFROST_MODE:-aggressive}"
Then explain to the user, in plain language:
- Whether the cache anchor (tools + system) is staying frozen or churning, and the count of resets.
- Any volatile content still detected in the system prefix (dates, git status, UUIDs, hashes) and what it costs.
- If
ENABLE_TOOL_SEARCHis nottruewhileANTHROPIC_BASE_URLis a custom host, warn that Claude Code is likely re-inlining the full tool set every turn (a major buster) and that it must be set before launchingclaude. - Concrete next steps (e.g. switch to
PERMAFROST_MODE=aggressive, set the env var, restart the session).
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 · 21 lines · 13 tokens per session scan A d84f4dd955e9
doctor is a command published in the GitHub repository jianzhichun/permafrost (22 stars, last pushed 2mo ago), licensed MIT. It adds 13 tokens to every session and 328 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
dev
Developer tools — debug, commit, PR, code review, refactoring.
verify
A command that independently checks the factual claims and file-and-line references in an agent’s most recent response. It reports whether each claim is confirmed, outdated, missing, or unsupported.
code-quality
Code review and bug prediction — find quality issues, style violations, and likely bugs.
fix-test
Auto-diagnose and fix failing tests — identifies root causes and applies fixes.
refactor
Code-level refactoring analysis and roadmap generation.
distil-doctor
Diagnose your distil setup — ledger, shadow validation, proxy round-trip, wiring.