---

name: brownfield description: "Scan and manage brownfield repository/worktree defaults for interviews"

/ouroboros:brownfield

Scan a root directory for existing git repositories and linked worktrees, then manage default repos used as context in interviews.

Usage

ooo brownfield                # Scan repos and set defaults
ooo brownfield scan           # Scan only (no default selection)
ooo brownfield defaults       # Show current defaults
ooo brownfield set 6,18,19   # Set defaults by repo numbers
ooo brownfield detect [path]  # Author mechanical.toml via one AI call

Trigger keywords: "brownfield", "scan repos", "default repos", "brownfield scan", "mechanical detect"

---

How It Works

Default flow (ooo brownfield with no args)

Step 1: Scan

Show scanning indicator:

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  Scanning for Existing Projects...
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Looking for git repositories and worktrees up to two directories below the scan root.
Local repos and repos with any remote name are eligible.
This may take a moment...

Implementation — use MCP tools only, do NOT use CLI or Python scripts:

CRITICAL — deferred-schema guard (prevents "Invalid tool parameters"): This skill can call ouroboros_brownfield across multiple turns (scan, set_defaults, defaults, and set). A deferred schema loaded for one turn is NOT guaranteed to remain loaded for the next. Immediately before EVERY ouroboros_brownfield call, re-run tool discovery query: "+ouroboros brownfield" (idempotent — a no-op when already loaded). If the load returns no matching tool (and the tool is not already callable — an empty load for an already-exposed tool is an expected no-op, not absence), stop with the MCP-not-available message instead of retrying the failing call.

1. Load the brownfield MCP tool: tool discovery query: "+ouroboros brownfield"
2. Call scan+register:

   Tool: ouroboros_brownfield
   Arguments: { "action": "scan" }
   

   This walks scan_root (up to two directory levels deep) for valid seed repos/worktrees and registers them in DB. Each repo or worktree found directly by the walk is registered self-only — Git worktree families are not expanded, so worktrees outside the depth-bounded walk (e.g. under .ouroboros/worktrees) are not pulled in. Existing defaults are preserved.

The scan response text already contains a pre-formatted numbered list with [default] markers. Do NOT make any additional MCP calls to list or query repos.

Display the repos in a plain-text 2-column grid (NOT a markdown table). Use a code block so columns align. Example:

Scan complete. 8 repositories registered.

 1. repo-alpha                   5. repo-epsilon
 2. repo-bravo *                 6. repo-foxtrot
 3. repo-charlie                 7. repo-golf *
 4. repo-delta                   8. repo-hotel

Include * markers for defaults exactly as they appear in the scan response.

If no repos found, show:

No git repositories or worktrees found.

Then stop.

Scan boundaries

• The filesystem walk starts at scan_root; when omitted, scan_root defaults to the current user's home directory.
• Repositories are discovered by walking directories inside scan_root, at most two levels deep (so ~/repo and ~/group/repo are found; deeper nesting is not).
• Dot-prefixed directories and known noisy directories such as node_modules are not walked as seed locations.
• Both normal repos (.git directory) and linked worktrees (.git file) are registered when the walk reaches them. Git worktree families are NOT expanded — a worktree is only registered if the walk finds it directly, not because its main repo's Git metadata reports it.
• Local repos, repos without remotes, and repos whose remotes are not named origin are all eligible.

Step 2: Default Selection

Do NOT use AskUserQuestion for this selection. Two rendering constraints make it unusable here:

• Assistant text emitted between tool calls (scan result → grid → question dialog) is not guaranteed to render, so the user may face the dialog without ever seeing the repo list.
• Embedding the list in option preview fields does not work either — the preview box has a fixed height and silently truncates long lists ("N lines hidden"), and repo lists routinely exceed it.

Instead, end the turn with the repo grid as the final message so its display is guaranteed, and collect the selection as a plain chat reply.

Immediately below the grid (same message, same code block or right after it), append the selection prompt:

If defaults exist:

Current defaults: <current default names> (numbers <current default numbers>)

Reply with repo numbers to change defaults (e.g. "6, 18, 19"),
"keep" to keep the current defaults, or "none" to clear them.

If no defaults exist:

No defaults set.

Reply with repo numbers to set defaults (e.g. "6, 18, 19"),
or "none" to run interviews in greenfield mode.

Then end the turn — no tool calls after the grid. The RFC #1392 breadcrumb footer is still the last line of the message.

On the next turn, parse the user's reply:

• Numbers (any separator) → those indices
• "keep" (defaults exist) → stop; no MCP call needed, confirm defaults unchanged
• "none" → empty indices (clear all)
• Anything else → ask again in plain text; do not guess

Then re-run tool discovery query: "+ouroboros brownfield" and use ONE MCP call to update all defaults at once:

Tool: ouroboros_brownfield
Arguments: { "action": "set_defaults", "indices": "<comma-separated IDs>" }

Example: if the user picks IDs 6, 18, 19 → { "action": "set_defaults", "indices": "6,18,19" }

This clears all existing defaults and sets the selected repos as default in one call.

If "none" → { "action": "set_defaults", "indices": "" } to clear all defaults.

Step 3: Confirmation

Brownfield defaults updated!
Defaults: grape, podo-app, podo-backend

These repos will be used as context in interviews.

Or if "none" selected:

No default repos set. Interviews will run in greenfield mode.
You can set defaults anytime with: ooo brownfield

Or if "keep" selected:

Defaults unchanged: <current default names>

---

Subcommand: scan

Scan only, no default selection prompt. Show the numbered list and stop.

---

Subcommand: defaults

Re-run tool discovery query: "+ouroboros brownfield", then call:

Tool: ouroboros_brownfield
Arguments: { "action": "scan" }

Display only the repos marked with * (defaults). If none, show:

No default repos set. Run 'ooo brownfield' to configure.

---

Subcommand: set <indices>

Directly set defaults without scanning. Parse the comma-separated indices from the user's input, re-run tool discovery query: "+ouroboros brownfield", and call:

Tool: ouroboros_brownfield
Arguments: { "action": "set_defaults", "indices": "<indices>" }

Show confirmation with updated defaults.

---

Subcommand: detect [path]

Runs one AI call against the target directory (defaults to the user's cwd) and writes .ouroboros/mechanical.toml with validated lint / build / test / static / coverage commands. Stage 1 of evaluation reads this file verbatim, so the toml is the authoritative Stage 1 contract — no hardcoded language presets exist anymore.

Ouroboros auto-runs this detect the first time ouroboros_evaluate is invoked without a toml present, so most users never need to call it directly. Run it explicitly when:

• you want to pre-author the toml before the first evaluate,
• you moved to a new build tool and want to refresh (--force),
• you want to review/edit the commands before Stage 1 trusts them.

Implementation: invoke the CLI via Bash.

uvx --python '>=3.12' --from ouroboros-ai ouroboros detect [path]
# or, if already installed:
ouroboros detect [path] [--force]

Then print the resulting .ouroboros/mechanical.toml contents so the user can confirm the proposed commands or hand-edit them.

If detect reports "could not propose any verifiable commands", surface the reason (no manifests found, LLM unavailable, every proposal dropped) and suggest the user write a minimal toml by hand — any single entry like test = "pytest -q" is enough to opt back in to Stage 1 for that check.

RFC #1392 State Breadcrumb Footer

Your final response MUST end with exactly one breadcrumb footer line:

◆ <current state> → next: <recommended action>

Derive <current state> from live session state via ouroboros_session_status when that MCP projection is available; otherwise derive it from this skill's actual outcome. Never use a linear Step N of M footer because Ouroboros is an evolutionary loop. When the next action is genuinely a choice, list 2-3 honest options in the next: clause. The breadcrumb line must be the last line of the response.