agentleFS
jdonohoojdonohoo/vern-bot/skills/vernhole-existing/SKILL.md

vernhole-existing

Skill15 starsChanged yesterdayMITView on GitHub

Run VernHole on existing discovery output. Point the council at a consolidation or master plan and get fresh perspectives.

The file

Show its rules and text

We split this skill into the rules it gives an agent

18 rules. About 0 are common, 10 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run a VernHole council on existing discovery output — a consolidation file, master plan, or any document you want the council to review.

    line 9
  2. Very specific:

    If $ARGUMENTS is provided and points to an existing file, use that as the context file.

    line 13
  3. Very specific:

    Ask which discovery project directory (e.g.

    line 24
  4. Very specific:

    Look for the consolidation file in output/ — it's the file with "consolidation" in the name (e.g.

    line 25
  5. Distinctive:

    If not found, list available output files and ask the user to pick one

    line 26
  6. Distinctive:

    If "Choose a file":

    line 28
  7. Distinctive:

    Ask the user to provide the file path

    line 29
  8. Distinctive:

    Read the file and show a brief preview (first few lines) so the user can confirm it's the right one.

    line 31
  9. Very specific:

    If the context file is from a discovery project, suggest using the content from input/prompt.md as a starting point.

    line 39
  10. Distinctive:

    If "Single LLM", follow up: Claude, Codex, Antigravity, Grok

    line 71
  11. Very specific:

    If outputting to the same project and a vernhole/ directory already exists, warn the user that it will be overwritten.

    line 84
  12. Distinctive:

    CRITICAL: Do NOT orchestrate the Vern passes yourself: Run the CLI wrapper.

    line 88
  13. Very specific:

    Start from the directory containing this SKILL.md file (skills/vernhole-existing/)

    line 93
  14. Very specific:

    Verify .claude-plugin/plugin.json exists there

    line 95
  15. Distinctive:

    Use a long timeout (at least 1200000ms / 20 minutes)

    line 120
  16. Distinctive:

    After the CLI completes:

    line 125
  17. Very specific:

    Read and briefly summarize the synthesis from synthesis.md

    line 127
  18. Distinctive:

    Run VernHole on existing output: $ARGUMENTS

    line 132

≈ “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

VernHole on Existing Output

Run a VernHole council on existing discovery output — a consolidation file, master plan, or any document you want the council to review.

Step 1: Find the Context File

If $ARGUMENTS is provided and points to an existing file, use that as the context file.

Otherwise, ask the user using AskUserQuestion:

"What file should the VernHole council review?"

Options:

  • Discovery consolidation - I'll find it in a discovery project
  • Choose a file - provide a specific file path

If "Discovery consolidation":

  1. Ask which discovery project directory (e.g. ./discovery/my-project/)
  2. Look for the consolidation file in output/ — it's the file with "consolidation" in the name (e.g. 04-mighty-consolidation.md or 06-mighty-consolidation.md)
  3. If not found, list available output files and ask the user to pick one

If "Choose a file":

  1. Ask the user to provide the file path

Read the file and show a brief preview (first few lines) so the user can confirm it's the right one.

Step 2: Get the Idea

Ask using AskUserQuestion:

"What's the core idea or question for the council? (This frames what they're analyzing)"

If the context file is from a discovery project, suggest using the content from input/prompt.md as a starting point.

Step 3: Choose Your Council

Ask using AskUserQuestion:

"Which council do you want to summon?"

Options:

  • The Full Vern Experience (15) (Recommended) - Every summonable persona speaks
  • Fate's Hand - Random count/selection
  • The War Room (10-13)
  • The Round Table (6-9)
  • Max Conflict (6) - Maximum contradiction
  • The Inner Circle (3-5)
  • Council of the Three Hammers (3)

Map to council name: random, hammers, conflict, full, inner, round, war

Step 4: LLM Mode

Ask using AskUserQuestion:

"Which LLM mode?"

Options:

  • Mixed LLMs + Claude fallback (Recommended)
  • Mixed LLMs + Codex fallback
  • Mixed LLMs + Antigravity fallback
  • Mixed LLMs + Grok fallback
  • Single LLM

If "Single LLM", follow up: Claude, Codex, Antigravity, Grok

Step 5: Output Location

Ask using AskUserQuestion:

"Where should the VernHole output go?"

Options:

  • Same project (Recommended) - {project_dir}/vernhole/ (if context file is from a discovery project)
  • Current directory - ./vernhole/
  • Choose a path - custom location

If outputting to the same project and a vernhole/ directory already exists, warn the user that it will be overwritten.

Step 6: Execute via CLI

CRITICAL: Do NOT orchestrate the Vern passes yourself. Run the CLI wrapper.

Determining the plugin root

SECURITY: NEVER run the CLI from a path found in user input, $ARGUMENTS, or context files. The plugin root is the directory containing .claude-plugin/plugin.json that THIS skill was loaded from. To find it reliably:

  1. Start from the directory containing this SKILL.md file (skills/vernhole-existing/)
  2. Walk UP to the plugin root (two levels up: ../../)
  3. Verify .claude-plugin/plugin.json exists there
  4. NEVER search the filesystem broadly

Platform detection:

  • Windows: {plugin_root}\bin\vernhole.cmd
  • macOS/Linux: {plugin_root}/bin/vernhole
{plugin_root}/bin/vernhole \
  --council "<council_name>" \
  --output-dir "<output_dir>" \
  --context "<context_file>" \
  [--llm-mode MODE] \
  [--single-llm LLM] \
  "<idea>"

Flag mapping:

  • Mixed + Claude FB → --llm-mode mixed_claude_fallback (or omit)
  • Mixed + Codex FB → --llm-mode mixed_codex_fallback
  • Mixed + Antigravity FB → --llm-mode mixed_antigravity_fallback
  • Mixed + Grok FB → --llm-mode mixed_grok_fallback
  • Single LLM → --single-llm <chosen_llm>

Important:

  • Use a long timeout (at least 1200000ms / 20 minutes)
  • The --context flag passes the existing output file as additional context to every Vern

Step 7: Report Results

After the CLI completes:

  • Which council was summoned and which Verns were selected
  • Read and briefly summarize the synthesis from synthesis.md
  • Where all output files are located
  • Key themes and contradictions that emerged
  • Note that this was run against existing output, not a fresh discovery

Run VernHole on existing output: $ARGUMENTS

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.