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 OutlineDriven/outline-driven-development --skill autoplangit clone --depth 1 https://github.com/OutlineDriven/outline-driven-developmentWrote 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/outlinedriven/outline-driven-development/autoplan)<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/autoplan"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/autoplan.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00047 | $0.00950 |
| Opus 5 | $0.00023 | $0.00475 |
| Sonnet 5 | $0.00009 | $0.00190 |
| Haiku 4.5 | $0.00005 | $0.00095 |
Grade A, and why
autoplan 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 yesterday.
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 — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Autoplan
Contract
| Field | Bound contract |
|---|---|
| Trigger | The user runs /autoplan on a plan or idea. |
| Authority | Reversible local: writes only the amended plan, the task JSONL ledger at tasks/autoplan-ledger.jsonl, and TODO updates; rollback is discarding those local artifacts. No remote mutation. |
| Side effect | Writes the amended plan, appends tasks to tasks/autoplan-ledger.jsonl, and updates TODOs. |
| Done | The final approval gate presents an amended plan and aggregated implementation tasks with stable IDs. |
Inputs
Required: a plan or idea to develop, supplied as a file path or inline text.
Optional: an existing tasks/autoplan-ledger.jsonl to extend. If absent, create one.
Ledger destination and overwrite policy
The task ledger lives at tasks/autoplan-ledger.jsonl. When the file does not exist, create it. When it exists, append new tasks to the end; never overwrite or truncate existing entries. Each run appends only the tasks it produces. This preserves prior runs and makes the ledger append-only across invocations.
Task ID generation
Each task gets a deterministic ID formed from the reviewing phase prefix and a zero-padded sequence number within that phase:
- CEO review:
CEO-01,CEO-02, ... - Design review:
DES-01,DES-02, ... - DX review:
DX-01,DX-02, ... - Engineering review:
ENG-01,ENG-02, ...
Sequence numbers start at 01 within each phase and increment in the order tasks are identified. IDs are stable within a run: the same plan evaluated the same way produces the same IDs. If a phase produces no tasks, its prefix is unused.
Each ledger line is a JSON object with id, description, and phase keys.
Procedure
- Read the supplied plan or idea. If it is a path, read the file; if inline text, use it directly. Stop if no plan or idea is supplied. Done when: the plan or idea is read or the absence is reported.
- Run the four review phases (CEO, design, DX, engineering), each evaluating the idea against a six-principle decision register (clarity, impact, audience, risk, sequencing, reversibility). Record each phase's decisions in the register. If a phase cannot reach a decision, record the open question and continue the remaining phases. Done when: all four phases have recorded their decisions or open questions in the register.
- Amend the plan with the consolidated decisions from all four phases. Done when: the plan is amended with consolidated decisions.
- Derive implementation tasks from the four phases. Assign deterministic IDs per the task ID generation rules above. Construct the task ledger by appending one JSON line per task to
tasks/autoplan-ledger.jsonl, creating the file if absent. Update TODOs to reflect the aggregated tasks. Done when: the ledger is appended with one task per line, each carrying a stable ID, description, and phase, and TODOs reflect the aggregated tasks. - Present the amended plan and aggregated tasks at a final approval gate. Stop and wait for the human decision; do not proceed to execution. Done when: the approval gate is presented and the skill stops for the human decision.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- yesterday Changed · -24 tokens per session 7a9cac991140
- 4d ago First seen · 58 lines · 71 tokens per session scan A 8f412841dedb
autoplan is a skill published in the GitHub repository OutlineDriven/outline-driven-development (52 stars, last pushed 2d ago), licensed Apache-2.0. It adds 47 tokens to every session and 950 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
drift-detect
Use when the user says "plan drift", asks whether the roadmap, plans, or docs still match the code, or is deciding what to rebuild when restarting a stalled project. For doc-vs-code drift inside a specific diff, use sync-docs.
github-backlog-triage
Use when the user invokes backlog triage for a GitHub repo's open issues and PRs. Not for proactive triage, non-GitHub trackers, or single bug issues: use github-bug-report-triage.
setup-repo-skills
Use when the user wants one-time repository setup for tracker, triage labels, and domain conventions. Not for ongoing triage, issue creation, or multi-repo setup.
git-history-analysis
Use when the user asks about recent engineering work, what the team is working on, planning or roadmap material, or an explicitly requested Slack summary. Not for remote or irreversible changes.
graph-backbone
Use when defining, revising, or gate-replanning the project structural backbone in project-root graph.yaml; produces a validated approved revision preserving immutable topology and additive metadata. Not for remote, credential, publish, deploy, or irreversible changes.
wontfix
Use when the user wants to elicit refused directions, generalize them, and close matching tracker items as not planned. Not for duplicates, spam, or items closed as completed or obsolete.