agentleFS
Sign inStart free
pickleshellpickleshell/next-level-agent/skills/next-level-agent/SKILL.md

next-level-agent

Skill7 starsChanged 5 months agoMITView on GitHub

Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 6 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If you were dispatched as a subagent to execute a specific task, ignore this skill.

    line 7
  2. Distinctive:

    If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

    line 11
  3. Distinctive:

    IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE.

    line 13
  4. Distinctive:

    Before entering plan mode: if you haven't already brainstormed, invoke the brainstorming skill first.

    line 22
  5. Distinctive:

    When multiple skills apply, process skills come first — they set the approach, then implementation skills (frontend-design, etc.) carry it …

    line 28
  6. Distinctive:

    If your harness appears here, read its reference file for special instructions:

    line 54
  7. Very specific:

    Use nla_state to replace the complete private workflow ledger after classification, user approval, a material decision, each milestone, a …

    line 78
  8. Very specific:

    Use test_writer before implementation when acceptance criteria can be tested through public contracts.

    line 105
  9. Very specific:

    When the plugin reports context pressure or continuity is at risk, stop starting large subagent tasks, save a complete ledger, and call …

    line 112
  10. Very specific:

    Do not emulate /compact. The tool saves the ledger immediately, then at the next safe idle boundary performs a Supervisor audit, creates a …

    line 114
  11. Very specific:

    After restoration, continue from next_step.

    line 119

≈ “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
<SUBAGENT-STOP> If you were dispatched as a subagent to execute a specific task, ignore this skill. </SUBAGENT-STOP><EXTREMELY-IMPORTANT> If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

This is not negotiable. You cannot rationalize your way out of this. </EXTREMELY-IMPORTANT>

The Rule

Invoke relevant or requested skills BEFORE any response or action — including clarifying questions, exploring the codebase, or checking files. If it turns out wrong for the situation, you don't have to use it.

Before entering plan mode: if you haven't already brainstormed, invoke the brainstorming skill first.

Then announce "Using [skill] to [purpose]" and follow the skill exactly. If it has a checklist, create a todo per item.

Skill Priority

When multiple skills apply, process skills come first — they set the approach, then implementation skills (frontend-design, etc.) carry it out. Brainstorming and systematic-debugging are common NLA process skills, but the rule holds for any of them.

  • "Let's build X" → brainstorming first, then implementation skills.
  • "Fix this bug" → systematic-debugging first, then domain skills.

Red Flags

These thoughts mean STOP—you're rationalizing:

ThoughtReality
"This is just a simple question"Questions are tasks. Check for skills.
"I need more context first"Skill check comes BEFORE clarifying questions.
"Let me explore the codebase first"Skills tell you HOW to explore. Check first.
"I can check git/files quickly"Files lack conversation context. Check for skills.
"Let me gather information first"Skills tell you HOW to gather information.
"This doesn't need a formal skill"If a skill exists, use it.
"I remember this skill"Skills evolve. Read current version.
"This doesn't count as a task"Action = task. Check for skills.
"The skill is overkill"Simple things become complex. Use it.
"I'll just do this one thing first"Check BEFORE doing anything.
"This feels productive"Undisciplined action wastes time. Skills prevent this.
"I know what that means"Knowing the concept ≠ using the skill. Invoke it.

Platform Adaptation

If your harness appears here, read its reference file for special instructions:

  • Codex: references/codex-tools.md
  • Pi: references/pi-tools.md
  • Antigravity: references/antigravity-tools.md
  • Hermes Agent: references/hermes-tools.md

NLA Coordination State

These rules apply only to the primary nla agent. Subagents must not read or write the shared notebook.

Fast durable memory

For substantive work, invoke assistant-notebook and use nla_notebook once at session entry, project switch, or after context restoration. Read Contents and only the relevant subject page. Reuse the recovered context; do not reread the notebook on every turn. Current user instructions and verified artifacts override notebook notes. Update only durable verified decisions, preferences, milestones, blockers, artifact references, and next steps. Never store secrets, transcripts, large logs, or speculative reasoning.

Session ledger

Use nla_state to replace the complete private workflow ledger after classification, user approval, a material decision, each milestone, a blocker, verification, and before completion. Preserve goal, Tier, workflow stage, acceptance criteria, approved decisions, completed and active work, changed files, verification evidence, blockers, pending gate, and exact next step.

Supervisor gates

Supervisor audits a bounded ledger and never edits files, dispatches agents, or accesses the notebook.

  • Tier 3: invoke Supervisor through nla_task at the design gate, execution gate, material milestones or anomalies, before compaction, and before completion.
  • Tier 2: invoke Supervisor only after repeated failure, BLOCKED, repeated NEEDS_CONTEXT, scope drift, context pressure, compaction, or a long-running milestone.
  • Tier 0 and Tier 1: do not invoke Supervisor.

NLA executes the verdict. Supervisor does not control the session directly.

Planner and Test Writer

For approved substantial work, use the planner role to turn requirements and architecture into a dependency graph with independently verifiable tasks and safe parallel groups. NLA remains responsible for dispatch and integration.

Use test_writer before implementation when acceptance criteria can be tested through public contracts. It writes only tests and test-owned fixtures and must produce valid RED evidence. The Implementer owns production code; Reviewer independently checks the resulting implementation and evidence.

Context compaction

When the plugin reports context pressure or continuity is at risk, stop starting large subagent tasks, save a complete ledger, and call nla_compact. Do not emulate /compact. The tool saves the ledger immediately, then at the next safe idle boundary performs a Supervisor audit, creates a Compactor checkpoint through the model pool, calls native OpenCode session summarization, and restores the validated checkpoint. This post-response lifecycle requires a persistent OpenCode TUI/server; one-shot opencode run may exit first. After restoration, continue from next_step; consult the notebook only if durable context is actually missing.

User Instructions

User instructions (CLAUDE.md, AGENTS.md, GEMINI.md, etc, direct requests) take precedence over skills, which in turn override default behavior. Only skip skill workflows or instructions when your human partner has explicitly told you to.

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.

Check your setup, free