hjemmesidekongen/plan-engine
plugins/kronen/skills/plan-engine/SKILL.md
Goal-driven iterative planning engine. OODA loop that plans one wave at a time, executes via multi-agent dispatch, and accumulates learnings across cycles. Always iterative, always goal-driven. Use when breaking work into waves, planning multi-file changes, or running structured execution.
development
Updated Apr 5, 2026
$ install --global
skillsauthnpx skillsauth add hjemmesidekongen/ai plan-engineInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 5, 2026, 5:22 PM349.1s5 files scanned
SKILL.md
- name:
- plan-engine
- description:
- >
- user_invocable:
- false
- interactive:
- false
- type:
- data_validation
- - name:
- verification_contract
- verify:
- Wave has concrete verification requirements in plan.yml
- fail_action:
- Generate verification contract before execution
- on_fail:
- Fix issues and re-run checkpoint
- on_pass:
- Report cycle number, classification, wave count, and active learnings count
- model_tier:
- principal
- origin:
- kronen
- inspired_by:
- Reflexion, AdaPlanner, GoalAct + plan-engine + dynamic-planner
- ported_date:
- 2026-03-13
- iteration:
- 1
- changes:
- Consolidated from plan-engine + dynamic-planner. Single iterative-only planning mode.
plan-engine
Goal-driven iterative planning. Plans one wave, executes, learns, loops. Goal fixed, path adapts.
OODA Cycle (7 phases)
- Orient — read learnings (active only), scan codebase, assess position vs goal. Cycle 1: run 6-point goal clarity gate (blocks if <4/6 pass). Cycle 2+: re-enforce, ESCALATE on regression.
- Reflect — dispatch plan-classifier agent for unbiased classification: CONTINUE / ADJUST / REPLAN / ESCALATE. Skip on cycle 1. Fallback: inline classification if agent fails.
- Research — optional. Triggered when reflect finds knowledge gaps. Scoped, time-boxed.
- Plan — decompose next wave: topological sort, file-ownership isolation, model-tier assignment, verification contract. ONE wave only.
- Build — dispatch task agents via context-manager. Parallel where file-ownership allows. Each agent gets: task spec, plan.md rules, dependency artifacts, verification contract.
- Verify — dispatch plan-verifier agent (isolated, no build context). Checks against upfront verification contract. Stage 1: mechanical. Stage 2: quality review.
- Learn — record typed entries (observation/constraint/correction/discovery) to learnings.yml. Error patterns become instinct entries. Update remaining_goal. Increment cycle.
Safety guards
- Goal clarity gate: 6-point mechanical checklist before cycle 1. Blocks on <4/6 pass.
- Goal immutability: goal + acceptance_criteria frozen after cycle 1. Changes → ESCALATE.
- Max 2 replans per goal → then ESCALATE
- max_cycles (default 15) → then ESCALATE
- ESCALATE always surfaces — never silently skipped
- Completed waves never undone on replan
Key rules
- Goal fixed at creation — never changes. Autopilot by default.
- ONE wave at a time. Only active learnings loaded. Verification contract upfront.
References
references/process.md— full algorithm. Load when executing the OODA loop.references/reflect-prompts.md— adversarial templates. Load during Reflect phase only.- Schemas:
plugins/kronen/resources/plan-schema.yml,plugins/kronen/resources/state-schema.yml,plugins/kronen/resources/learnings-schema.yml
Related Skills
hjemmesidekongen/brand-strategy
development
VerifiedTrustedCommunity
Creates a brand from scratch through market research and interactive sparring. Runs competitive research via Perplexity, then guides the user through positioning, audience, voice, values, and content pillars. Produces the full brand guideline set at .ai/brand/{name}/. Use when building a new brand, defining brand strategy for a product, or when /våbenskjold:create is invoked.
SKILL.mdUpdated Apr 5, 2026
hjemmesidekongen/brand-loader
testing
VerifiedTrustedCommunity
Loads brand guidelines from .ai/brand/{name}/ and makes them available to the current context. Progressive disclosure: L1 confirms brand exists, L2 loads summary, L3 loads specific files on demand. Use when a downstream skill or user needs brand context, or when /våbenskjold:apply is invoked.
SKILL.mdUpdated Apr 5, 2026
hjemmesidekongen/brand-evolve
documentation
VerifiedTrustedCommunity
Guided reinvention of an existing brand guideline. Loads current brand from .ai/brand/{name}/, identifies what to keep vs change, and walks the user through targeted evolution. Preserves brand equity while updating positioning, voice, or values. Use when refreshing a brand or when /våbenskjold:evolve is invoked.
SKILL.mdUpdated Apr 5, 2026
hjemmesidekongen/brand-audit
development
VerifiedTrustedCommunity
Codifies an existing brand from materials, samples, and references. Analyzes provided content to extract voice patterns, values, and positioning. Produces the same guideline format as brand-strategy. Use when a brand already exists but isn't documented, or when /våbenskjold:audit is invoked.
SKILL.mdUpdated Apr 5, 2026