agentleFS
p10ns11yp10ns11y/skills/master-planner/SKILL.md

master-planner

Skill2 starsChanged 40 days agoMITView on GitHub

Builds a project master plan as an interconnected web of friction-removing automations aimed at impossible missions. Resolves the skills library from SKILLS_ROOT, ~/Work/personal/skills, or clones https://github.com/p10ns11y/skills when local is missing; pulls a pack into cwd, Orwell-tweaks overlays, and adds ontology for large/complex repos. Use for master plans, skill pull/tweak, project skill packs, ontology, or /master-planner.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read library README.md pack table.

    line 81
  2. Very specific:

    Keep in-repo skills (not symlinks) when they encode project-only architecture (e.g.

    line 97
  3. Distinctive:

    Create when ≥2 of: multi-crate/control-plane, agent-facing contracts, easy wrong-file edits, cross-module invariants.

    line 121
  4. Very specific:

    Update Active skills table: skill · path · when.

    line 135

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Master Planner

Load rule: Formal SoT below. Orwell/ontology depth → references/ only if needed.
CLT: install ../rules/clt-dual-load.mdc with packs (A8 DualLoad pre-filter); ontology/overlays must reduce rediscovery load, not add second SoTs.

// Mission
MasterPlan ≔ friction-removing automation web for an impossible mission (live skill web — not slides)

// Resolve
SKILLS_ROOT → ~/Work/personal/skills → clone https://github.com/p10ns11y/skills
Project ──symlink──► .agents/skills/ | .cursor/skills/
Portable bodies: never fork — overlays only
Project-born skills: keep in-tree; publish via [dual-copy-skill-publish](../dual-copy-skill-publish/SKILL.md)

// Axioms
A1  Plan ⇔ removes friction (tokens | handoffs | rediscovery | unsafe defaults)
A2  Symlink portable skills; overlay tweaks; never dual-edit portable SKILL.md
A3  Ontology only when complex (≥2 of: multi-crate, agent contracts, wrong-file risk, cross-module invariants)
A4  Verify pack + project checks before "done"

Remote: p10ns11y/skills. Companions: ai-optimization · fusion-sage · higher-order-decision-architect · stellar-spacemap · agent-orchestrator


When to run

SignalAction
“Master plan”, “skill pack”, “pull skills”, /master-plannerFull workflow
New repo / cold agent setupPull + AGENTS.md + verify
Large complex repo, agents lose the mapAdd/refresh ontology
Overlays stale vs shipped architectureTweak overlays only

Skip: one-file typo, single bugfix with known path.


Workflow (do in order)

- [ ] 1. Orient — cwd, AGENTS.md, existing .agents/skills, mission one-liner
- [ ] 2. Scan library — list skills; map pack to stack (see Pack router)
- [ ] 3. Pull — symlink missing skills; keep in-repo skills that are project-owned
- [ ] 4. Tweak — Orwell + action + diagram + compress overlays (see Tweak law)
- [ ] 5. Ontology — if complex: INDEX + graph (+ GRAPH.md); else skip
- [ ] 6. Wire — AGENTS.md active-skills table + verify block
- [ ] 7. Verify — script + project checks; report gaps

1. Orient

pwd
test -f AGENTS.md && head -80 AGENTS.md
ls -la .agents/skills 2>/dev/null || ls -la .cursor/skills 2>/dev/null

Capture: mission (one sentence) · stack signals (Cargo.toml, install.sh, package.json, …) · already wired skills.

2. Scan library

# Resolves local or clones https://github.com/p10ns11y/skills
SKILLS_ROOT="$(./scripts/resolve-skills-root.sh)"
ls "$SKILLS_ROOT" | grep -vE '^(examples|rules|\.git|LICENSE|README)'

Read library README.md pack table. Prefer starter packs over one-off picks.

3. Pull

# Resolves SKILLS_ROOT (local → remote clone) then symlinks
./scripts/pull-skills.sh --project "$(pwd)" --pack <pack-name>
./scripts/pull-skills.sh --project "$(pwd)" --skills ai-optimization,fusion-sage,...
# force remote even if a stale local path exists:
./scripts/pull-skills.sh --project "$(pwd)" --pack arch-guardian --from-remote

Rules:

  • Symlink into .agents/skills/<name> (arch-machine style) or .cursor/skills/ — match repo convention.
  • Never edit portable SKILL.md bodies for one project — use .agents/overlays/ or examples/overlays/.
  • Keep in-repo skills (not symlinks) when they encode project-only architecture (e.g. eagle TEA).

4. Tweak (overlays)

Load references/orwell-tweak.md. For each pulled skill that needs local coords:

WritePath
Relevance / never-compress.agents/overlays/<project>-ai-optimization.md
Fused aggregates.agents/overlays/<project>-fusion-sage.md
Decision zones.agents/overlays/<project>-decision-hooks.md
Roadmap scout.agents/overlays/<project>-stellar-spacemap.md
Master pack map.agents/overlays/<project>-master-planner.md

Tweak law (non-negotiable):

  1. Orwell — short words; cut dead metaphors; prefer active verbs; one idea per line.
  2. Action — every section ends in a verb the agent can run (path, command, refuse).
  3. Diagram — one mermaid or ASCII arch for the fused loop; drop second prose copy.
  4. Compress — tables > paragraphs; drop legacy unless still live; no token waste.
  5. Specific — real paths and verify commands for this cwd.

5. Ontology (large / complex only)

Create when ≥2 of: multi-crate/control-plane, agent-facing contracts, easy wrong-file edits, cross-module invariants.

Assets (template: references/ontology-template.md):

FileRole
.agents/ontology/INDEX.mdIntent → subgraph router
.agents/ontology/<project>.graph.yamlNodes + edges + source_refs
.agents/ontology/GRAPH.mdOptional mermaid viz

Point agents: load subgraph by intent; never dump full graph.

6. Wire AGENTS.md

Update Active skills table: skill · path · when. Point overlays. Keep verify block exact.

7. Verify

./scripts/verify-pack.sh --project "$(pwd)"
# then project checks from AGENTS.md (lint, tests, --validate, …)

Done when:

  • Pack symlinks resolve
  • Overlays match live architecture (no dead entrypoints)
  • Ontology (if any) subgraphs route without contradiction
  • AGENTS.md lists active skills + verify commands
  • Report: pack · mission · friction removed · next automation

Pack router

Stack signalsPackCore skills
install.sh + profiles + maintenance + archy/TUIarch-guardianmaster-planner, eagle-*, ai-optimization, fusion-sage, HODA, stellar-spacemap, verification-cockpit, session-unit-order, agent-orchestrator, control-graph, git-worktrees
~/.config/shell / path.contractshell-verifyshell-kernel-ontology, verification-cockpit, stellar-spacemap, ai-optimization, fusion-sage
Next.js / React appweb-appai-optimization, fusion-sage, react-client-expert, semantic-markup-css, fix-dependency-security
Tauri + agent loopsagentic-desktopfinder-reactor, tauri-agentic, agent-orchestrator, git-worktrees, control-graph
Multi-agent delivery onlymulti-agentagent-orchestrator, control-graph, git-worktrees, concurrent-cli-agents, split-to-prs
Strategy / backlog docsstrategyHODA, stellar-spacemap, fusion-sage, master-planner, odysseus-navigator

Custom: compose from catalog; document in overlay.


Master plan output (when user wants the plan, not only the pull)

## Master plan — <project>

**Mission:** <one sentence — the impossible made routine>

**Friction web:**
| Automation | Removes | Skill / path | Verify |
|------------|---------|--------------|--------|

**Arch:**
\`\`\`mermaid
flowchart LR
  ...
\`\`\`

**Refuse / build:** …
**Next wave:** …

Pair roadmap depth with stellar-spacemap. Pair material bets with higher-order-decision-architect.


Anti-patterns

¬Do
prose roadmap, no automation edgesfriction web table + skills
fork portable SKILL.md per projectoverlays + symlink
pull whole library “just in case”pack router
ontology = AGENTS.md copyaddressable IDs + subgraphs
dead shims as primaryupdate or delete overlays
funeral closename acceleration automation

Refs: orwell-tweak · ontology-template · arch-machine-pack · master-planner.mdc

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.