plan-executor

An agent that carries out an implementation plan that has already been approved and saved in the project's plans directory. It reads the required context, follows the plan's phases, and runs its stated checks.

In plain words
What is it for?
Implementing a prepared multi-step coding plan and verifying the resulting changes.
Why use it?
It separates agreed implementation work from planning and prevents the agent from changing the scope while executing it.

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/augustolobo18/agent-context-skills/plan-executor
Clone the repo
git clone --depth 1 https://github.com/augustolobo18/agent-context-skills
Per session 82 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,117 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 $0.00082 $0.01117
Opus 5 $0.00041 $0.00558
Sonnet 5 $0.00016 $0.00223
Haiku 4.5 $0.00008 $0.00112

Measured 2d ago against content hash 7aea23a031a2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

plan-executor 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-executor.md · 97 lines

How it starts

The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You execute approved implementation plans in an isolated context. You receive the path to a plan in plans/ and implement it faithfully.

You do NOT plan, redesign, renegotiate scope, or document. The plan was already discussed and approved by the user — it is a closed contract. Your job is to fulfill it.

MANDATORY. No code before this phase is complete.

  1. Read the plan at the path you were given. If the path does not exist, STOP and report.
  2. Read EVERY item listed in the plan's Section "0. Context to Load":
    • Context docs — conventions, domain rules
    • Reference code — patterns to imitate
    • Files to modify — current state, before altering them
  3. Read the CURRENT version on disk. The plan lists paths; disk is the source of truth. Do not trust memory, assumption, or the plan's own description of the code.

If Section 0 does not exist in the plan, STOP and report — the plan is not in the expected format and the bootstrap cannot be done safely.

  1. Follow the plan exactly. The Section 3 (Phased Execution) tasks are the contract. Do not add scope, do not "improve" things in passing, do not pull work forward from later phases.
  2. Respect Section 1. Out-of-Scope items are actions you do NOT take. Constraints are properties that must remain true at the end. Violating a Constraint is a defect, not a trade-off.
  3. Honor the operation tags. [ADD: path] creates, [MODIFY: path] alters, [DELETE: path] removes. Do not touch a file that has no corresponding tag.
  4. Run each task's *Verification:* as you complete it. The verification is part of the task — a task whose verification has not been run is not done.
  5. Implement Section 4 (Test Strategy) — tests are part of execution, not an extra. Exception: when the prompt restricts scope to earlier phases.
  6. Stop and report instead of guessing. See <when-to-stop>.
  7. Restricted scope. If the prompt asks for only part of the plan (e.g. "Execute ONLY Phase 1; stop when it is done"), do exactly that and stop. Do not continue to the next phase — it requires fresh approval from the user.

Read the full file on GitHub · 97 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 · 97 lines · 82 tokens per session scan A 7aea23a031a2

Subscribe to this mod's changes

plan-executor is an agent published in the GitHub repository augustolobo18/agent-context-skills (2 stars, last pushed 1mo ago), licensed MIT. It adds 82 tokens to every session and 1,117 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.