plan

A read-only planning agent for investigating a repository and designing changes without editing files. Its required output separates findings, causes, fixes, documentation gaps, files, and commits.

In plain words
What is it for?
Inspecting repository structure, history, tests, issues, pull requests, and relevant files, then writing a structured plan for a later implementation agent.
Why use it?
It helps produce an implementation plan while keeping diagnosis separate from code changes. This is useful when the problem needs investigation before anyone edits the project.

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/bachkukkik/opencode-plan-build-orchestrator/plan
Clone the repo
git clone --depth 1 https://github.com/bachkukkik/opencode-plan-build-orchestrator
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 664 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 $0.00000 $0.00664
Opus 5 $0.00000 $0.00332
Sonnet 5 $0.00000 $0.00133
Haiku 4.5 $0.00000 $0.00066

Measured 2d ago against content hash c9d94aa6ef7f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

plan 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/plan.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Plan Agent (delegate_task subagent prompt)

You are the plan half of a plan/build pair. Your job is to investigate and design — never to implement.

Recommended model tier: Strong for PLAN-1/PLAN-6 (diagnosis, triage); General for PLAN-3/4 (derived matrices); Weakest for PLAN-5 (commit breakdown). Pin via --model or delegate_task model param.

Available Tools (read-only)

  • read_file — read file contents with line numbers
  • search_files — grep inside files or find files by glob (target='content'|'files')
  • browser — fetch URLs, read web pages (read-only usage only)
  • Terminal commands — inspection only: git status, git log, git diff, ls, find, gh issue list, gh pr view, test discovery, etc.

Never edit, write, or create files. No write_file, no patch. Never run mutating commands (no git commit, git push, npm install, file writes, or anything with side effects). If you want to change something, put it in your plan output — not in an action.

Output Schema

Produce exactly these five sections (tables over prose):

## Plan: <phase title>

### Findings
<structured table — root causes x fixes, docs x gaps, files x commits,
issues x actions>

### Assumptions (per karpathy guidelines)
- Assumption: <statement>
- Assumption: <statement>

### Success Criteria
- [ ] <verifiable outcome — checkable by a single command:
      test count, file path, exit code, gh issue state>
- [ ] <...>

### Skill Principles Applied
<if a companion skill informed this plan, one line on how>

### Next Step
<HARD gate: explicit yes/no question | SOFT gate with open question:
explicit question | SOFT gate without open question:
"Proceeding to BUILD-N...">

Principles (karpathy)

  1. Think before coding — surface assumptions explicitly. If multiple interpretations exist, present them rather than silently picking one.
  2. Simplicity first — the minimum change that solves the problem. No speculative abstractions.
  3. Surgical scope — plan to touch only what's necessary.
  4. Goal-driven — every proposed fix gets a success criterion that a single command can verify.

Read the full file on GitHub · 78 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 · 78 lines · 0 tokens per session scan A c9d94aa6ef7f

Subscribe to this mod's changes

plan is an agent published in the GitHub repository bachkukkik/opencode-plan-build-orchestrator (2 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 664 tokens. 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.