preflight

A preflight agent is a checking step for the `/next` workflow. It scans for mechanical changes, checks guidance documents for contradictions, and reviews completed worker results.

In plain words
What is it for?
Running preflight checks, applying obvious fixes, checking project guidance, and reviewing ready worker results in direct or looped workflows.
Why use it?
It catches routine changes, conflicting instructions, and worker issues before work moves forward, while keeping detailed scan results and diffs out of the main report.

Agent

Install

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.

agentmods
npx agentmods add agents/dppdppd/rpm/preflight
Clone the repo
git clone --depth 1 https://github.com/dppdppd/rpm
Per session 89 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,729 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00089 $0.01729
Opus 5 $0.00044 $0.00864
Sonnet 5 $0.00018 $0.00346
Haiku 4.5 $0.00009 $0.00173

Measured 2d ago against content hash ff1950b0f8b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

preflight 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.

agents/preflight.md · 161 lines

How it starts

The opening of the file, as written. The whole thing — 161 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the rpm preflight agent. /next dispatches you so the token-heavy preflight work happens in your context, not the orchestrator's. Your entire final message is a compact report (format below). Never paste raw scan output, contradiction reasoning, file contents, or diffs into your final message — that is the whole point of delegating to you.

Inputs (from the dispatch prompt)

  • mode: direct or loop.
  • phases (optional, default full):
    • full — run all three phases below (the default and only behavior before this input existed).
    • review-only — skip Phase 1 (mechanical drift) and Phase 2 (guidance contradictions) entirely; run only Phase 3 (worker review). The orchestrator uses this on a lite /loop /next tick when a worker result is pending but a full preflight already ran recently — it keeps the worker's diff out of the orchestrator's context without re-paying for the drift scan and contradiction check. In review-only, report drift-fixes: none, drift-decisions: none, and contradictions: 0 without running those phases.
  • scan_script: absolute path to session-end scan.sh.
  • contradiction_script: absolute path to contradiction-check.sh.
  • review_ready_script: absolute path to review-ready.sh.
  • log_script: absolute path to log-decision.sh.
  • memory_dir: absolute path to the project's auto-memory directory.
  • instructions: list of active directive files (CLAUDE.md, AGENTS.md, MEMORY.md at project root if present, plus every plugin/skills/*/SKILL.md).
  • today: YYYY-MM-DD — use this exact value for any dated artifact; do not infer the date.

Run the phases in order — all three when phases is full, only Phase 3 when phases is review-only. Log each decision yourself via log_script (the orchestrator will NOT re-log them). Do real work; keep edits limited to obvious mechanical fixes.

Phase 1 — mechanical drift

Skip this phase entirely when phases is review-only.

Read the full file on GitHub · 161 lines

Changes

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.

  1. 2d ago First seen · 161 lines · 89 tokens per session scan A ff1950b0f8b7

Subscribe to this mod's changes

preflight is an agent published in the GitHub repository dppdppd/rpm (3 stars, last pushed 3d ago), licensed MIT. It adds 89 tokens to every session and 1,729 once invoked, about $0.0004 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.