claude-roundtable

name

claude-roundtable

description

Bring routed frontier models into the Claude Code harness for consult, critique, or implementation via the local CLIProxyAPI proxy — currently nicknamed sol, terra, and luna (fetch /v1/models if the list may have changed; more providers like Gemini/Grok/Kimi show up automatically once logged in). No slash command required — trigger on natural language like "ask terra", "with sol", "use luna", or any name that isn't a Claude model. Typos are fine; match against the live model list. /consult, /critique, /implement exist as explicit shortcuts but are not required.

Claude Roundtable — consult / critique / implement on routed models

You are the ORCHESTRATOR, running on this session's normal Anthropic login. NEVER set ANTHROPIC_BASE_URL / ANTHROPIC_AUTH_TOKEN on this session's own environment — routed env vars go inline on individual headless calls only. (Anthropic's terms ban routing a Claude subscription login through a proxy; this flow never does that.)

For Claude-to-Claude mixing (Fable ↔ Sonnet ↔ Opus) no routing is needed: the user can /model mid-session. This skill is for the OTHER subscriptions.

Config

Env var	Default	Meaning
MODEL_ROUTER_URL	http://127.0.0.1:8317	proxy address
MODEL_ROUTER_KEY	my-proxy-key	key from the proxy's config.yaml
MODEL_ROUTER_MODEL	gpt-5.6-sol	default routed model
MODEL_ROUTER_EFFORT	high	effort for IMPLEMENT only

CONSULT and CRITIQUE always run at high effort (the remote model is doing the thinking). IMPLEMENT runs at MODEL_ROUTER_EFFORT (the spec carries the thinking).

Model picking (fuzzy, forgiving): the user speaks casually — "with terra", "ask sol", "use luna" — and may typo ("dera", "tera"). Map whatever they said to the closest ID in the proxy's /v1/models list (e.g. "terra" → gpt-5.6-terra). Fetch the list if you haven't this session. Always state which model you picked ("Routing to gpt-5.6-terra") BEFORE the call, so a wrong guess is visible immediately. If nothing plausibly matches, use the default and say so. If the user names no model, use MODEL_ROUTER_MODEL (gpt-5.6-sol) without comment.

Preflight — once per session, before the first routed call

curl -s "${MODEL_ROUTER_URL:-http://127.0.0.1:8317}/v1/models" \
  -H "Authorization: Bearer ${MODEL_ROUTER_KEY:-my-proxy-key}"

Connection refused → start the proxy (~/cli-proxy-api/cli-proxy-api.exe -config ~/cli-proxy-api/config.yaml, backgrounded) and re-check. Auth error or empty model list → STOP and tell the user their provider login needs (re)running: cd ~/cli-proxy-api && ./cli-proxy-api.exe -config config.yaml -codex-login.

CONSULT — a second opinion (read-only)

ANTHROPIC_BASE_URL="${MODEL_ROUTER_URL:-http://127.0.0.1:8317}" \
ANTHROPIC_AUTH_TOKEN="${MODEL_ROUTER_KEY:-my-proxy-key}" \
claude -p "<question, with relevant file/dir PATHS for it to read>" \
  --model "${MODEL_ROUTER_MODEL:-gpt-5.6-sol}" --effort high \
  --bare --max-turns 15 --output-format json \
  --disallowedTools "Edit,Write,NotebookEdit"

Prompt rules:

• Pass PATHS, never file contents — the worker reads them on its own quota.
• End every consult prompt with: "Return your verdict and top points in 30 lines or fewer."
• The JSON output includes a session_id — keep it for follow-ups.

CRITIQUE — same call as CONSULT, different framing

Prompt shape: "Critique : strengths, top 5 issues ranked by severity, one concrete fix each. 40 lines or fewer."

FOLLOW-UP — cross-model memory (the important trick)

The remote model remembers its own session. To continue a consult/critique (e.g. hand it Fable's counterpoint), resume — send ONLY the new delta:

ANTHROPIC_BASE_URL="${MODEL_ROUTER_URL:-http://127.0.0.1:8317}" \
ANTHROPIC_AUTH_TOKEN="${MODEL_ROUTER_KEY:-my-proxy-key}" \
claude -p --resume <session_id> "<only the new information or counterpoint>" \
  --model "${MODEL_ROUTER_MODEL:-gpt-5.6-sol}" --effort high \
  --bare --max-turns 15 --output-format json \
  --disallowedTools "Edit,Write,NotebookEdit"

Never re-paste what the remote model already saw.

IMPLEMENT — build via a routed model (one compound call per task)

ANTHROPIC_BASE_URL="${MODEL_ROUTER_URL:-http://127.0.0.1:8317}" \
ANTHROPIC_AUTH_TOKEN="${MODEL_ROUTER_KEY:-my-proxy-key}" \
claude -p "<self-contained task spec>" \
  --model "${MODEL_ROUTER_MODEL:-gpt-5.6-sol}" \
  --effort "${MODEL_ROUTER_EFFORT:-high}" \
  --bare --max-turns 30 --permission-mode acceptEdits \
&& ls -la <expected output paths> \
&& grep -c "<acceptance marker>" <built file>

The spec must be self-contained: exact file paths, what to create or change, acceptance criteria, what NOT to touch, and "verify your work, then report changed paths + how you verified". You make every design call in the spec — the worker just types. Misses go back as a fix delegation with the failure evidence attached; you do not fix code yourself.

TOKEN RULES (non-negotiable — this is why the skill exists)

1. Paths, never contents, in delegate prompts.
2. Cap every response ("≤30 lines") — whatever comes back lives in this session's context forever.
3. Independent opinions run in PARALLEL (your own analysis in-session while the routed call runs); chain Fable→Sol only when one must critique the other.
4. Follow-ups go through --resume, sending only the delta.
5. Never read a whole built file back into context — verify with targeted grep -c / head / tail slices.