iterative-planner

State-machine driven iterative planning and execution for complex coding tasks. Cycle: Explore → Plan → Execute → Reflect → Pivot. Filesystem as persistent memory. Use for multi-file tasks, migrations, refactoring, failed tasks, or anything non-trivial.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A sophisticated, highly actionable state-machine protocol with strong workflow sequencing, validation gates, and excellent progressive disclosure into the bundled references and scripts. Its main weakness is conciseness: at ~400 lines it repeats some lifecycle/ownership and compression material that could be consolidated.

Suggestions

Consolidate the overlapping file-access information: the File Lifecycle Matrix and the File Ownership Model table cover the same files from two angles — merge or cross-reference once to cut ~30-40 lines.

Describe consolidated-file compression once (Consolidated File Management) and have the Lifecycle Matrix footnote point to it, rather than restating the intra-plan compression trigger inline and again in the Lessons section.

Tighten the Orchestrator Role Assumption section, which is the longest single block; the three-mode dispatch could be expressed as a compact table or checklist without the repeated reload-loop rationale.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes Claude's competence (e.g. 'Context Window = RAM. Filesystem = Disk.') and aggressively delegates per-state detail to emit-state.mjs modules and references, but at ~400 lines it repeats material — the File Lifecycle Matrix and File Ownership Model overlap, and consolidated-file compression is described twice — that could be tightened.	2 / 3
Actionability	Concrete executable commands throughout: 'node <skill-path>/scripts/bootstrap.mjs new "goal"', 'node <skill-path>/scripts/emit-state.mjs --state explore', 'bootstrap.mjs retire <plan-id>', and 'git checkout <checkpoint-commit> -- .' are copy-paste ready.	3 / 3
Workflow Clarity	The state machine is explicitly sequenced via a mermaid diagram plus a transitions table, with hard validation checkpoints — 'validate-plan.mjs --pre-step' (exit 2, GATE:FAIL slugs), the 7 mandatory REFLECT Gate-In reads, and the revert-first Autonomy Leash feedback loop.	3 / 3
Progressive Disclosure	SKILL.md is an overview spine with well-signaled, one-level-deep references to verified bundle files (references/.md, scripts/modules/state-.md); per-state detail and templates are appropriately split out, making navigation easy.	3 / 3
	Total	11 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A tight, well-constructed description that covers concrete capabilities, natural trigger terms, both what and when, and a distinctive niche in third-person voice. The only soft spot is the breadth of 'anything non-trivial', which slightly weakens distinctiveness.

Dimension	Reasoning	Score
Specificity	Names multiple concrete actions — 'state-machine driven iterative planning and execution', 'Filesystem as persistent memory', and the explicit five-state 'Explore → Plan → Execute → Reflect → Pivot' cycle.	3 / 3
Completeness	Explicitly answers both what it does ('state-machine driven iterative planning and execution ... Filesystem as persistent memory') and when to use it ('Use for multi-file tasks, migrations, refactoring, failed tasks, or anything non-trivial').	3 / 3
Trigger Term Quality	'multi-file tasks, migrations, refactoring, failed tasks, or anything non-trivial' are natural phrases a user would say when they need this skill, with good coverage of common variations.	3 / 3
Distinctiveness Conflict Risk	The state-machine framing, named five-state cycle, and filesystem-as-memory hook carve a distinct niche unlikely to fire for ordinary single-edit coding, though 'anything non-trivial' is somewhat broad.	3 / 3
	Total	12 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 1 missing, 6 deeper-than-1-level	Warning

	Total	14 / 16 Passed

Repository: NikolasMarkou/iterative-planner
Commit: 219878c

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.