roadmap

A roadmap manager for the APE runtime. It stores durable requirements and derives their statuses from runtime history.

In plain words
What is it for?
Use it to check roadmap status, register approved requirements, supersede existing entries, and record attestations.
Why use it?
It keeps roadmap requirements recorded and prevents invalid dependency graphs or unapproved changes from being accepted.

Skill for Claude CodeCodex

Part of the ape plugin — 7 skills, 1 MCP server shipped together

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 skills/aawwcc/ape/roadmap
Any agent
npx skills add AAWWCC/ape --skill roadmap
Clone the repo
git clone --depth 1 https://github.com/AAWWCC/ape

Made for: Claude Code, Codex.

Or install ape, the plugin that ships this one along with the rest of its 7 skills, 1 MCP server.

Per session 16 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 633 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 $0.00016 $0.00633
Opus 5 $0.00008 $0.00316
Sonnet 5 $0.00003 $0.00127
Haiku 4.5 $0.00002 $0.00063

Measured 2d ago against content hash ff799906a812, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

roadmap 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.

plugins/ape-claude/skills/roadmap/SKILL.md · 44 lines

What it actually says

APE roadmap

Use only when the user explicitly requests roadmap work. The roadmap stores durable requirements; its statuses are always derived by the runtime, never supplied by the caller. On Google Antigravity / Gemini, pass the exact open project root as project_dir.

  • status: call ape_history with action: "roadmap-status". Optional status_filter narrows returned entries without changing aggregate counts. roadmap: null means no roadmap exists.
  • register: first draft cold-reader-complete entries and obtain explicit approval. Then call roadmap-register with a non-empty audit reason and at most 64 entries. Each entry needs id, title, description, and observable acceptance; depends_on and discovered_by are optional. Same-batch forward references are allowed. Unknown or stale dependencies, duplicate edges, self-reference, cycles, or an otherwise-invalid live graph reject the entire batch.
  • supersede: only after explicit approval, call roadmap-supersede with selected IDs and a non-empty reason. Targets and optional replacements must be unique, known, live, and disjoint. Supersession marks entries stale; it does not rewrite history.
  • attest: close live requirements against an archived completed run or the exact verified produce-and-hold shape (blocked at merge, gates passed, canonical auto-merge-disabled reason). Call roadmap-attest with requirement_ids, run_id, and a non-empty audit reason; requirements must be known and live. No other blocked shape is eligible, and a hold does not satisfy from completes alone. Use when work was merged outside APE, a run completed before the roadmap existed, or fully gated work awaits shipping by policy.

Never pass a status field. Registration is atomic. A proposed entry must describe a behavioral or operator consequence if omitted; style notes and isolated documentation nits remain advisory rather than becoming standalone requirements.

When a run discovers follow-up work, declare normalized proposals in the accepted receipt's evidence.roadmap_followups array, without status or discovered_by, then ask for approval before registration. Receipt acceptance does not register anything. Use the discovering run ID only when that active or archived run contains an accepted receipt with the exact declaration; the runtime rejects nonexistent runs and payload mismatches.

Roadmap mutations are journaled and exact-once. A divergent journal/store state is a hard refusal, not permission to retry with changed input. A run targeting a roadmap entry starts and completes only while every dependency is satisfied; stale targets and all other dependency states block.

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 · 44 lines · 16 tokens per session scan A ff799906a812

Subscribe to this mod's changes

roadmap is a skill published in the GitHub repository AAWWCC/ape (0 stars, last pushed 3d ago), licensed MIT. It adds 16 tokens to every session and 633 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 2026-08-31.