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 commands/punt-labs/prfaq/externalizegit clone --depth 1 https://github.com/punt-labs/prfaqWhat 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 | $0.00017 | $0.01440 |
| Opus 5 | $0.00009 | $0.00720 |
| Sonnet 5 | $0.00003 | $0.00288 |
| Haiku 4.5 | $0.00002 | $0.00144 |
Grade A, and why
externalize 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 — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Externalize: PR/FAQ → Press Release
Convert the internal PR/FAQ decision document into a customer-facing press release for a specific release. The PR/FAQ is what the team debates; the press release is what the world reads.
Steps
-
Find the inputs. Read
prfaq.tex(or the path the user specifies) andCHANGELOG.mdfrom the project directory. Both are required — the PR/FAQ provides the narrative and the CHANGELOG provides what actually shipped. If either is missing, tell the user what's needed and stop. -
Determine the target version. Three branches based on
$ARGUMENTS:a) Version specified —
$ARGUMENTScontains a version string (e.g.,v2.0,2.0,0.9.1): target that version's CHANGELOG entry. If the version doesn't exist in the CHANGELOG, tell the user and list available versions.b) Empty — default to the latest version in the CHANGELOG (the topmost
## [X.Y.Z]entry that is not[Unreleased]). If the CHANGELOG has no versioned releases (only[Unreleased]), tell the user that externalize requires at least one versioned release and stop. -
Detect the release type. Read the full CHANGELOG to understand the product's release history. The release type determines the tone, structure, and length of the press release:
Release type Condition Structure First release Target is the first (or only) version in the CHANGELOG Full press release: headline, lede, problem, solution, customer quote, getting started, spokesperson quote, CTA, boilerplate Major update Target has a major version bump from the prior release (e.g., v1.x → v2.0) and prior versions exist Update announcement: headline focuses on "what's new", lede highlights why this upgrade matters, "What's New" section replaces "Problem", solution focused on new capabilities, CTA includes upgrade path Minor/patch Target is a minor or patch bump (e.g., v1.0 → v1.1 or v1.0.1) Short release note: headline, lede summarizing improvements, key changes list, availability, boilerplate
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 · 66 lines · 17 tokens per session scan A fc83f5f8b9fc
externalize is a command published in the GitHub repository punt-labs/prfaq (25 stars, last pushed 2d ago), licensed MIT. It adds 17 tokens to every session and 1,440 once invoked, about $0.0001 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 2026-08-30.
Other commands, from other repositories
oracle-dev
Generate property-based test harness using Lean 4 model as oracle.
refine-dev
Verify code refines Z specification via abstraction function and commutativity checks.
refine
Verify code refines Z specification via abstraction function and commutativity checks.
contracts-dev
Generate runtime contracts (preconditions, postconditions, invariants) from a Z specification.
prove-dev
Generate Lean 4 proof obligations from a Z specification.
prove
Generate Lean 4 proof obligations from a Z specification.