openplan-align

A guided method for clarifying an ambiguous request before research, design, implementation, or review.

In plain words
What is it for?
Use it to restate a request, separate facts from assumptions, identify decisions and blind spots, and choose the next appropriate kind of work.
Why use it?
It helps expose missing goals, users, boundaries, risks, and success criteria before an agent takes the wrong route.

Skill for Claude CodeCodex

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/bananasjim/openplan-skill/openplan-align
Any agent
npx skills add BANANASJIM/openplan-skill --skill openplan-align
Clone the repo
git clone --depth 1 https://github.com/BANANASJIM/openplan-skill

Made for: Claude Code, Codex.

Per session 78 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 516 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 $0.00078 $0.00516
Opus 5 $0.00039 $0.00258
Sonnet 5 $0.00016 $0.00103
Haiku 4.5 $0.00008 $0.00052

Measured 2d ago against content hash 24dc7bae5e3f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

openplan-align 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.

skills/openplan-align/SKILL.md · 56 lines

What it actually says

OpenPlan Align

Use this after $openplan-core when the next safe action depends on understanding the goal/intent.

Alignment Flow

  1. Restate the user's request in one paragraph.
  2. Separate facts from assumptions.
  3. Walk the 7-layer funnel only as far as needed.
  4. Identify blind spots and human decisions.
  5. Recommend the next route: ask via the human-facing coordinator, research, design, implement, review, or stop.

7-Layer Funnel

Layer Question Enough When
L1 WHY Why does this need to exist? Job-to-be-done is explicit
L2 WHO/WHERE Who uses it and in what constraints? Context and stakeholders are clear
L3 WHAT What is in and out of scope? Boundaries are explicit
L4 HOW macro What modules or artifacts are involved? Interfaces and ownership are named
L5 RISK What might fail or be unknown? Risks have a research or mitigation path
L6 VERIFY How will success be checked? Testable criteria exist
L7 DETAIL What exact implementation detail matters? Work is actionable

Do not force all seven layers for a small request. Use the smallest depth that prevents a wrong next action.

Output

For lightweight alignment:

Goal/Intent:
Known facts:
Assumptions:
Blind spots:
Human decisions:
Recommended next route:

For formal alignment, use references/alignment-contract.md.

Rules

  • When the user gives a solution, identify what problem it solves unless already clear.
  • Ask at most 1-3 questions only if this agent is the human-facing coordinator; otherwise list Human decisions required.
  • Do not make the human's tradeoff decision.
  • If enough is clear, proceed with a stated assumption instead of blocking.
  • If evidence is missing, recommend research rather than inventing facts.

For Claude Code, use assets/claude-code-command.md as a slash command or prompt seed.

Files

What ships with it

3 files 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.

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 · 56 lines · 78 tokens per session scan A 24dc7bae5e3f

Subscribe to this mod's changes

openplan-align is a skill published in the GitHub repository BANANASJIM/openplan-skill (4 stars, last pushed 3mo ago), licensed MIT. It adds 78 tokens to every session and 516 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.