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 entropy-assisted-planninggit 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/entropy-assisted-planning)<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/entropy-assisted-planning"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/entropy-assisted-planning/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/skills/outlinedriven/outline-driven-development/entropy-assisted-planning"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/entropy-assisted-planning.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00048 | $0.01425 |
| Opus 5 | $0.00024 | $0.00713 |
| Sonnet 5 | $0.00010 | $0.00285 |
| Haiku 4.5 | $0.00005 | $0.00143 |
Grade A, and why
entropy-assisted-planning 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 5d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- entropy-assisted-planning — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Entropy assisted planning
Contract
| Field | Bound contract |
|---|---|
| Trigger | The user explicitly requests a Tarot draw or casually delegates an ambiguous choice among multiple valid approaches. |
| Authority | Run a local cryptographically random draw and record the chosen direction in chat. No file, VCS, credential, paid, published, deployed, or remote mutation. Rollback is to discard the reading or redraw; no persistent state requires recovery. |
| Side effect | Chat output only: a 12-house draw and its interpreted direction. The draw command touches no repository or remote state. |
| Done | A complete 12-house reading yields a stated direction or verdict, and any security or correctness implication remains subject to ordinary evidence. |
Inputs
- Decision context (required): either a list of two or more valid options or an open portent question about which direction to take.
- User tone (optional): when the user seeks precision rather than a casual choice, do not draw; ask clarifying questions instead.
Procedure
- Confirm the trigger. Draw only when the user explicitly requested a Tarot draw or casually delegated an ambiguous choice among two or more valid approaches. If the user gave clear, specific instructions, or a single obvious correct approach exists, do not draw. Done when: the trigger is confirmed or the run is declined with a reason.
- Do not use this skill as the deciding authority for safety-critical work: security conclusions, data integrity, production deployment, release approval, or incident response. It may suggest where to inspect next but cannot prove safety or dismiss a risk. Done when: safety-critical scope is excluded or the reading is bounded to investigation-path suggestion.
- Run the cryptographic draw in one Bash call. The command uses Python's
secretsmodule. It shuffles separate 22-card Major Arcana and 56-card Minor Arcana decks with Fisher-Yates viasecrets.randbelow()(no modulo bias), deals 12 houses (one Major plus two Minor each), and gives each of the 36 cards an independent 50% reversal viasecrets.randbits(1). The conservative unordered-card entropy budget exceeds 100 bits.
Done when: the draw command exits zero and emits valid JSON with 12 houses.python3 - <<'PY' import secrets, json MAJOR=[f"{i:02d}-{n}" for i,n in enumerate(("the-fool","the-magician","the-high-priestess","the-empress","the-emperor","the-hierophant","the-lovers","the-chariot","strength","the-hermit","wheel-of-fortune","justice","the-hanged-man","death","temperance","the-devil","the-tower","the-star","the-moon","the-sun","judgement","the-world"))] RANKS=("ace","two","three","four","five","six","seven","eight","nine","ten","page","knight","queen","king") SUITS=("wands","cups","swords","pentacles") MINOR=[f"{r}-of-{s}" for s in SUITS for r in RANKS] HOUSES=("Self","Resources","Communication","Foundations","Creativity","Practice","Partnership","Transformation","Exploration","Calling","Community","The Hidden") def shuf(d): for i in range(len(d)-1,0,-1): j=secrets.randbelow(i+1); d[i],d[j]=d[j],d[i] return d shuf(MAJOR); shuf(MINOR) out=[] for i in range(12): out.append({"house":i+1,"name":HOUSES[i],"major":MAJOR[i],"minor1":MINOR[2*i],"minor2":MINOR[2*i+1],"reversed":[secrets.randbits(1)==1 for _ in range(3)]}) print(json.dumps(out,indent=2)) PY - Interpret the full 12-house spread as one narrative. Reversed cards invert or complicate the upright meaning rather than meaning "bad". Major Arcana carry more weight than Minor Arcana. Do not interpret any card in isolation; synthesize across all 12 houses before deciding. Done when: the 12-house spread is synthesized into one narrative.
- Map the interpretation to the concrete decision. If the input was a list of options, pick one option and state it with one sentence connecting card meaning to the choice. If the input was an open portent question, state the dominant theme across the spread, the main risk or blind spot, and the recommended next action. Done when: the interpretation is mapped to one chosen option or a three-bullet reading.
- In security, audit, or correctness contexts, the reading only chooses an investigation path or hypothesis to test. Accept or dismiss any risk only with ordinary engineering evidence: source review, tests, proofs, traces, reproduction, or exploitability analysis. A favorable card is never permission to ship, suppress a finding, skip validation, or overrule a concrete risk. Done when: any security or correctness implication is explicitly marked as subject to ordinary evidence.
- One draw per decision point; accept the reading. Do not redraw until a preferred result appears. Done when: the reading is accepted with no redraw for preference.
- Include the interpretation alongside the next action that implements the chosen direction. Do not output the interpretation as a text-only turn. Done when: the interpretation and the next action are both in the output.
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.
- 5d ago First seen · 61 lines · 48 tokens per session scan A b691cc892a06
entropy-assisted-planning is a skill published in the GitHub repository OutlineDriven/outline-driven-development (52 stars, last pushed 3d ago), licensed Apache-2.0. It adds 48 tokens to every session and 1,425 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
audit-project
Run an iterative multi-agent code audit until critical and high findings are resolved. Use when the user says "audit my code", "find all the bugs", "deep code audit", "iterative review", or "review until clean".
duet
Use when the user invokes /duet, says "pair with me", or faces aesthetic, architectural, or irreversible decisions.
goal-prompt-drafting
Use when asked to draft copy-ready /goal objectives for long-running agents; returns one normalized one-line objective with measurable end state, grounded proof, easy-out invariants, a stop clause, and a Missing list. Not for source or remote-system changes.
handoff-prompt
Use when the user asks for a handoff, delegation, or clipboard-ready prompt for another agent: a standalone path-free prompt copied to the clipboard, confirmed by title. Not for session-snapshot briefs — use handoff; never remote, credential, publish, deploy, or irreversible.
publish-branch
Use when asked to publish the checked-out branch: commit and push it on whatever branch it is, the default branch included. Not for creating branches, PRs, force pushes, or pushing any other branch; when the request excludes the default branch, use commit-push-current.
snippet-image-rendering
Use when the user explicitly names snipgrapher and wants code rendered to a polished PNG, SVG, or WebP at an explicit local path. Probes the installed CLI for supported flags and writes only the named image file. Not for other renderers, publishing, or remote actions.