Skip to content
Skills Library
AI & agents · Planning

agent-plan-execute-replan

/agent-plan-execute-replan

Separate planning from execution and replan when reality diverges, so a long task does not follow a plan that stopped being true. Use when work has many steps and early assumptions are likely to be wrong.

Repository
AI-SKILLS ↗
Licence
MIT
Last repository change
05.09.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 05.09.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Separate planning from execution and replan when reality diverges, so a long task does not follow a plan that stopped being true. Use when work has many steps and early assumptions are likely to be wrong.

Original author description

Separate planning from execution and replan when reality diverges, so a long task does not follow a plan that stopped being true. Use when work has many steps and early assumptions are likely to be wrong.

How it works

  1. Method
  2. Boundaries
Original SKILL.md headings · 38 lines, 265 words
SKILL.md preview
---
name: agent-plan-execute-replan
description: Separate planning from execution and replan when reality diverges, so a long task does not follow a plan that stopped being true. Use when work has many steps and early assumptions are likely to be wrong.
---

# Plan, execute, replan

An agent that plans once and executes blindly follows a stale plan into
the ground. One that replans constantly never finishes. The working
pattern is an explicit plan, execution against it, and replanning
triggered by divergence rather than by every observation.

## Method

1. **Plan in steps with observable outcomes.** Each step states what
   done looks like, which is what makes divergence detectable at all.
2. **Separate the planner from the executor.** The planner holds the
   goal and the sequence; executors do one step with only what that step
   needs (see agent-context-isolation).
3. **Check the outcome after every step.** Comparing what happened to
   what was expected is the entire mechanism; without it you have a
   script, not a loop.
4. **Replan on divergence, not on every step.** A step that succeeded as
   expected needs no replanning, and replanning by default burns context
   and loses coherence.
5. **Keep the goal fixed while the plan changes.** Drifting goals are
   how long-running agents end up solving a different problem than the
   one asked (see goal-driven-execution).
6. **Cap total steps and replans.** A hard ceiling turns an infinite
   loop into a reported failure, which is always the better outcome.
7. **Surface the plan to the human.** A visible plan is reviewable
   before expensive execution and is the cheapest intervention point
   (see agent-human-checkpoint).

## Boundaries

This pattern handles uncertainty; it cannot rescue an impossible or
under-specified goal, which needs a human to reframe. Every replan
costs a full planning pass. Steps with irreversible effects must be
gated regardless of how confident the plan is (see
agent-delegation-protocol).

Versions

The database has no other skills named agent-plan-execute-replan.

Database snapshot: 13.11.2026 · package version 59ee2341eba7 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100dos-replan-loop
    Run bounded, recurring `dos-replan` sweeps, optionally guarded by the workspace's release policy. Use when the planning queue needs unattended refresh for a fixed cadence rather than one manual sweep.
  • 100plan-prompts
    Use when asked to "create a development plan", "generate execution prompts", "replan a milestone", "update DEVELOPMENT_PLAN.md", "start M5", or "adapt future milestones" after predecessor work changes the current repository design. Not for discovering an uncertain destination or…
  • 100workflow-plan
    Unified planning skill - 4-phase planning workflow, plan verification, and interactive replanning. Triggers on \"workflow-plan\", \"workflow-plan-verify\", \"workflow:replan\".
  • 100replan
    Review and validate an existing implementation plan by launching parallel subagents that check it from multiple angles — codebase alignment, best practices, code standards, feasibility, and fresh perspectives. Use whenever you have a plan ready and want to validate it before…
  • 100dos-dispatch-loop
    Run recurring `dos-dispatch` cycles, switching to `dos-replan` when the backlog drains and stopping on the kernel's loop verdict. Use for unattended dispatch->replan->dispatch work across disjoint lanes.
  • 100dos-dispatch-loop
    Run recurring `dos-dispatch` cycles, switching to `dos-replan` when the backlog drains and stopping on the kernel's loop verdict. Use for unattended dispatch->replan->dispatch work across disjoint lanes.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.