CtrlK
BlogDocsLog inGet started
Tessl Logo

pantheon-ai/implementation-planner

Converts a PRD or requirements document into a structured, phased implementation plan with individual phase files and granular per-task files written to .context/plans/. Also restructures existing monolithic planning documents into digestible, hierarchical directory structures. Creates a root plan index summarising all phases, a numbered phase file per phase, and a numbered task file per task inside each phase directory.

92

3.25x
Quality

93%

Does it follow best practices?

Impact

91%

3.25x

Average score across 5 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

Evaluation results

99%

95%

Task: Create an implementation plan for a URL Shortener service

create-new-plan-from-prd

Criteria
Without context
With context

plan-root-created

0%

100%

root-readme-is-index-only

0%

100%

phases-directory-structure

0%

100%

phase-readme-has-goal-gate-deps

0%

100%

gate-is-concrete-not-vague

0%

100%

tasks-directory-exists-per-phase

0%

100%

task-identifier-format-correct

0%

100%

task-files-have-verification-section

0%

100%

task-verification-not-vague

0%

100%

tasks-scoped-to-single-file

0%

100%

slug-format-correct

25%

100%

scripts-used-for-scaffolding

0%

100%

validate-plan-script-run

0%

100%

completion-summary-reported

0%

75%

additive-no-deletion

100%

100%

reasonable-phase-count

0%

100%

95%

20%

Task: Restructure a flat planning document into a navigable hierarchy

restructure-monolithic-plan-into-hierarchy

Criteria
Without context
With context

output-under-correct-path

100%

100%

each-phase-has-directory

100%

100%

no-numeric-only-directory-names

100%

100%

no-generic-vague-names

100%

62%

every-directory-has-readme

100%

100%

leaf-files-have-required-sections

20%

100%

flatten-vs-subdivide-heuristics-applied

100%

75%

numbering-prefix-grouping-respected

100%

100%

source-not-deleted-before-validation

50%

100%

validate-structure-script-run

0%

100%

readme-links-resolve

100%

100%

max-depth-respected

100%

100%

activities-or-steps-subdirectory

0%

100%

readme-minimum-content

100%

100%

kebab-case-names

100%

100%

97%

90%

Task: Create an implementation plan for an E-commerce Checkout Redesign

naming-conventions-slug-and-identifier-compliance

Criteria
Without context
With context

plan-slug-is-kebab-case

0%

100%

phase-slugs-are-kebab-case

0%

100%

task-identifier-format-all-files

0%

100%

task-numbering-starts-at-01

0%

100%

task-numbering-sequential-no-gaps

0%

100%

phase-numbering-sequential-no-gaps

0%

100%

no-spaces-or-underscores-in-slugs

66%

100%

phase-and-task-zero-padding-two-digits

0%

100%

task-slug-after-identifier-is-descriptive

0%

100%

root-readme-is-index

0%

100%

phase-readmes-include-goal-gate

0%

100%

validate-plan-run

0%

100%

completion-summary-includes-paths

20%

40%

slug-length-reasonable

100%

100%

70%

64%

Task: Create an implementation plan for a large-scale SaaS platform

large-scope-phase-count-guardrail

Criteria
Without context
With context

scope-analysis-performed

25%

100%

user-asked-before-exceeding-8-phases

0%

0%

question-is-specific-not-generic

0%

100%

no-plan-created-without-answer

0%

100%

plan-within-8-phases-if-user-not-asked

0%

0%

root-readme-is-navigation-index

0%

100%

phase-readmes-have-gates

0%

100%

task-identifiers-correct-format

0%

100%

task-verification-sections-present

0%

100%

scripts-used-for-scaffolding

0%

100%

validate-plan-run-if-files-created

0%

100%

slug-format-correct

100%

100%

97%

45%

Task: Add a new phase to an existing implementation plan

task-isolation-and-verification-quality

Criteria
Without context
With context

phase-03-directory-created

100%

100%

existing-phases-untouched

100%

62%

phase-readme-has-gate

0%

100%

task-count-matches-scope

100%

100%

task-ids-continue-sequence

0%

100%

task-scoped-to-single-handler

87%

100%

task-verification-is-runnable

0%

100%

no-vague-verification

62%

100%

dependencies-declared

100%

100%

scaffold-scripts-used

0%

100%

validate-plan-run

0%

100%

root-readme-updated

100%

100%

slug-format-correct

100%

100%

completion-summary-reported

50%

100%

Evaluated
Agent
Claude
Model
Claude Sonnet 4.5

Table of Contents