LLM Council on Fireworks AI
dair-ai/dair-academy-plugins
Has several open-weight models answer a question, rank each other's anonymized answers, then lets a chairman model write the final response through Fireworks AI.
Stands up the Weave model router in Docker Compose and drives it with claude -p against a real or mocked upstream to reproduce and verify routing and streaming behavior.
$ npx skills add weave-os/router --skill test-claude-locally -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install weave-os/router test-claude-locally --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/test-claude-locally .claude/skills/test-claude-locally && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .claude/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locallyType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add weave-os/router --skill test-claude-locally -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install weave-os/router test-claude-locally --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.agents/skills/test-claude-locally .agents/skills/test-claude-locally && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .agents/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add weave-os/router --skill test-claude-locally -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install weave-os/router test-claude-locally --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.agents/skills/test-claude-locally .cursor/skills/test-claude-locally && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .cursor/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/weave-os/router.git --path .agents/skills/test-claude-locally--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add weave-os/router --skill test-claude-locally -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install weave-os/router test-claude-locally --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.agents/skills/test-claude-locally .gemini/skills/test-claude-locally && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .gemini/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install weave-os/router test-claude-locallyInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add weave-os/router --skill test-claude-locally -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .github/skills && cp -r skills-src/.agents/skills/test-claude-locally .github/skills/test-claude-locally && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .github/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add weave-os/router --skill test-claude-locally -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install weave-os/router test-claude-locally --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/weave-os/router.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.agents/skills/test-claude-locally .opencode/skills/test-claude-locally && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "test-claude-locally" agent skill from https://github.com/weave-os/router/tree/main/.agents/skills/test-claude-locally into .opencode/skills/test-claude-locally/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "test-claude-locally", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
test-claude-locallyStands up the Weave model router in Docker Compose and drives it with claude -p against a real or mocked upstream to reproduce and verify routing and streaming behavior.
This is the interactive counterpart to an automated pre-merge smoke test: it brings up the router in docker compose, points a one-off claude -p session at it through a --settings file rather than environment variables, and reads the local server logs to confirm behavior for a specific upstream model such as GLM-5.1, DeepSeek or Qwen. Two upstream modes are supported, a real provider API that needs working credentials and credits, or a deterministic mock upstream emitting an exact SSE shape with no credits required.
Several gotchas are called out up front: ~/.claude/settings.json's env block overrides shell environment variables, so setting ANTHROPIC_BASE_URL in the shell silently fails and the request still reaches production unless --settings <file> is used, and the user's own global Claude Code config must never be edited for this; claude -p is stateless across invocations, so a /force-model directive must be the first line of the same prompt as the actual task, not a separate call.
The router also ignores the request's model field entirely and routes through its own cluster scorer, so /force-model through a Claude Code session is the only way to pin a model, not a raw curl call; and the monorepo's pubsub emulator may already hold host port 8085, worked around with a docker-compose.override.yml that drops the router's host binding while still reaching the emulator over the compose network.
7 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 37d585b. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships 1 file in scripts/ (Python), which the agent can run.
Shell commands in SKILL.md call:
dockerclaudepython3curlmakepsqlgitFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use docker, curl and git, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Names these keys or tokens, usually read from environment variables:
FIREWORKS_API_KEYANTHROPIC_API_KEYFrom names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Weave Router Local Testing loads about 3.1k tokens when it runs. Until then it costs about 119 tokens; SKILL.md has 1,210 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check noted patterns worth knowing about, such as sudo or a known installer.
al provider** — set the provider key in `.env.local` (e.g. `FIREWORKS_API_KEY=...`) and restart `docker compose up -d secomes from `ROUTER_CLUSTER_VERSION` in `.env.local`; it may differ from prod, which is why `/force-model` (not the scorAutomated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
The full file from weave-os/router at commit 37d585b, republished under its Apache-2.0 licence (© weave-os). 1,210 words, ~3,074 tokens.
.claude/skills/test-claude-locally/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.For an automated pre-merge regression net (fixture-driven, asserts caching/streaming/decision-headers against real Anthropic), run
make smoke— see docs/SMOKE.md. This skill is the interactive counterpart: stand the stack up by hand and drive it withclaude -pto reproduce or verify a one-off bug.
Stand up the router in docker compose, point a one-off claude -p session at it via --settings, and read the local server logs to confirm behavior. Two upstream modes: the real provider API (needs a working key + credits) or a mock upstream that emits an exact SSE shape (deterministic, no credits).
~/.claude/settings.json env overrides inherited env vars. Setting ANTHROPIC_BASE_URL in the shell does NOT redirect claude — settings.json wins and the request silently goes to prod. Always redirect with claude --settings <file> (see scripts/local-settings.json). Never edit the user's global ~/.claude.json or ~/.claude/settings.json — that breaks their live session.claude -p is stateless across invocations. A standalone /force-model call does not persist to the next claude -p. Put /force-model <model> as the first line of the SAME prompt that contains the task.model field and routes via the cluster scorer. The ONLY way to pin a specific model is /force-model through a Claude Code session (raw curl cannot).docker-compose.override.yml (see workflow). The server still reaches the emulator over the compose network.- [ ] 1. Bring up the stack (handle port 8085)
- [ ] 2. Seed an API key
- [ ] 3. Choose upstream: real provider OR mock
- [ ] 4. Write a one-off local-settings.json
- [ ] 5. Drive with `claude -p --settings`, forcing the target model
- [ ] 6. Read local logs to verify behavior
- [ ] 7. Clean upcd <router-repo>
# Drop the pubsub host-port binding to avoid an 8085 conflict:
cat > docker-compose.override.yml <<'EOF'
services:
pubsub-emulator:
ports: !reset []
EOF
docker compose up -d --build server # --build picks up code changes
until curl -sf http://localhost:8080/health >/dev/null; do sleep 2; doneThe override file is gitignored-by-intent scaffolding — delete it in cleanup.
docker compose run --rm seedCopy the rk_... key it prints.
Real provider — set the provider key in .env.local (e.g. FIREWORKS_API_KEY=...) and restart docker compose up -d server. Confirm the boot log shows <Provider> provider enabled with the real base_url. Use this to confirm a model genuinely produces the behavior.
Mock upstream — for a deterministic, credit-free repro of a precise SSE shape. Point the provider's base URL at a local mock and restart:
python3 scripts/mock_openai_upstream.py >/tmp/mock.log 2>&1 & # serves :8099
# In docker-compose.override.yml under `server:`, add:
# environment:
# FIREWORKS_BASE_URL: http://host.docker.internal:8099/v1
# FIREWORKS_API_KEY: sk-mock
# extra_hosts: ["host.docker.internal:host-gateway"]
docker compose up -d serverEdit the mock's emitted chunks to match the upstream shape you're reproducing. Provider→env-var names live in internal/providers/provider.go; base-URL overrides are read in cmd/router/main.go (<PROVIDER>_BASE_URL).
cat > /tmp/local-settings.json <<EOF
{ "env": {
"ANTHROPIC_BASE_URL": "http://localhost:8080",
"ANTHROPIC_CUSTOM_HEADERS": "X-Weave-Router-Key: rk_REPLACE_ME"
}}
EOFcd <scratch-dir-with-files-to-act-on>
env -u CLAUDE_CODE_SESSION_ID -u ANTHROPIC_BASE_URL -u ANTHROPIC_CUSTOM_HEADERS \
claude -p 'First send exactly: /force-model z-ai/glm-5.1
Then <task that requires tool use>, then stop.' \
--settings /tmp/local-settings.json --max-turns 10 --verboseThe decision + completion log per action (one ProxyMessages complete) carries the signals you need. Strip ANSI first:
docker compose logs server --since=3m 2>&1 | sed -E 's/\x1b\[[0-9;]*m//g' \
| grep 'ProxyMessages complete' | grep 'decision_model=z-ai/glm-5.1'Useful fields: decision_model, decision_provider, upstream_finish_reason, suppressed_tool_calls, text_only_turn_nudged, tool_use_blocks, resp_stop_reason, stop_reason_demoted. Also grep for recovery nudge, tool-call loop, no-progress.
Confirm the model was actually served before drawing conclusions:
docker compose logs server --since=3m 2>&1 | sed -E 's/\x1b\[[0-9;]*m//g' \
| grep 'ProxyMessages complete' | grep -oE 'decision_model=[^ ]+' | sort | uniq -cIf you only see other models, /force-model didn't take — recheck step 5.
pkill -f mock_openai_upstream.py 2>/dev/null
rm -f docker-compose.override.yml /tmp/local-settings.json
# `docker compose down` if you want to stop the stackFor fixes to internal/billing + internal/server/middleware/balance_check.go + the usage-bypass path (internal/proxy/usage_bypass.go), the setup differs from the model-behavior workflow above:
ROUTER_DEPLOYMENT_MODE=managed on the server service. Billing auto-enables when the router schema's billing tables exist (they do after migrate), logged as Router billing enabled. In default selfhosted mode WithBalanceCheck is never wired and none of this triggers.ANTHROPIC_BASE_URL override (unlike Fireworks/Together/etc.) — cmd/router/main.go hardcodes anthropic.DefaultBaseURL. To point Anthropic at a mock, temporarily edit that line to config.GetOr("ANTHROPIC_BASE_URL", anthropic.DefaultBaseURL), rebuild, and revert it after testing (don't ship it in the fix PR). The boot log line still prints DefaultBaseURL cosmetically — verify the mock is actually hit via the mock's own logs, not the boot log.max_tokens, no tools) classify as Probe/Classifier/TitleGen — all hard-pinned, short-circuiting before the usage-bypass and scorer branches. A trivial "say hi" will wrongly hit the refusal path. Use a realistic turn: tools present + max_tokens>=4096 + a normal user message. Confirm turn_type=main_loop in the turnloop classified log before trusting the result.curl on the production router-key path — no claude binary needed, and it lets you pin the requested model (the usage-bypass path serves the requested model verbatim, so no /force-model needed here):curl -sS -N -D /tmp/h.txt -X POST http://localhost:8080/v1/messages \
-H "Authorization: Bearer $RK" \
-H "X-Weave-Anthropic-Subscription: sk-ant-oat01-anything" \
-H "anthropic-version: 2023-06-01" -H "Content-Type: application/json" \
-d '{"model":"claude-sonnet-4-5","max_tokens":4096,"stream":true,"tools":[{"name":"Bash","input_schema":{"type":"object"}}],"messages":[{"role":"user","content":"<realistic multi-sentence task>"}]}'sk-ant-oat... token is enough for router-side classification; the mock doesn't validate it.psql -h localhost -p 5433 -U router -d router, password router, SET search_path TO router): UPDATE model_router_installations SET usage_bypass_enabled=true WHERE external_id='__router_admin__'; and upsert organization_credit_balance with balance_usd_micros below SubscriptionOverdraftFloorMicros (−5_000_000). Assert no debit via SELECT count(*) FROM organization_credit_ledger WHERE organization_id='__router_admin__'; before/after.model in the request body is reliable).X-Router-Decision: usage_bypass, the depleted-credits warning injected as content-block index 0 (real content re-indexed to 1), log Balance past subscription overdraft floor: serving subscription-only. Retryable upstream error → log Subscription-only bypass hit retryable error; refusing instead of paid reroute → HTTP 402. No subscription header below floor → insufficient_credits 402 that never logs ProxyMessages start.ROUTER_CLUSTER_VERSION in .env.local; it may differ from prod, which is why /force-model (not the scorer) is the reliable way to hit one model.internal/router/catalog/catalog.go.git ls-tree <monorepo-commit> router-internal/router shows the pinned router submodule SHA.WV_CAPTURE_CONTENT=full on the Compose server and leave the seeded installation's content_capture_mode NULL. Deployment off wins over installation full; installation off wins over deployment full.router.model_router_installations.content_capture_mode do not invalidate cached auth. Restart the server after a direct update, or use the admin endpoint that publishes invalidation.x-weave-force-model: <catalog model> pins the target without a slash-command exchange. Supply it alongside X-Weave-Router-Key through the one-off Claude settings. --settings accepts inline JSON as well as a file, allowing locally seeded keys to remain in shell memory.last_preview can therefore show environment text rather than the task; title-generation requests may contain the task canary. Add a raw authenticated Messages request when exact preview equality matters.{"type":"object","properties":{"canary":{"type":"not_a_valid_json_schema_type"}}}. Verify upstream_status=400 and a nonempty error body with capture on before asserting it is blank off. Unknown model pins can fail locally, invalid tool names are sanitized, and invalid caller API keys are ignored on the router-keyed paid-credential path.ANTHROPIC_API_KEY for real Anthropic checks. Pass it through the process environment to Compose; do not copy it into tracked files or plaintext test artifacts.© weave-os, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 1 other file (scripts) in .agents/skills/test-claude-locally of weave-os/router.
Open the folder on GitHubat commit 37d585b
Weave Router Local Testing next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Weave Router Local Testing this skillweave-os/router | 5.6k | — | ~3.1k | Automated safety check: Notes | Apache-2.0 | |
| LLM Council on Fireworks AIdair-ai/dair-academy-plugins | 614 | — | ~5k | Automated safety check: Notes | MIT | |
| slime RL Post-TrainingOrchestra-Research/AI-Research-SKILLs | 13k | 4 repos | ~2.8k | Automated safety check: Pass | MIT | |
| Deepseek Harness Dockerrunzhliu/deepseek-harness-docker | 110 | — | ~2.7k | Automated safety check: Notes | MIT | |
| Cloud TestFreakStudioCN/mpy-hardware-extension | 154 | — | ~1.7k | Automated safety check: Pass | Custom licence | |
| Cliproxy Newapi Stackmajiayu000/spellbook | 287 | — | ~1.7k | Automated safety check: Warn | MIT |
dair-ai/dair-academy-plugins
Has several open-weight models answer a question, rank each other's anonymized answers, then lets a chairman model write the final response through Fireworks AI.
Orchestra-Research/AI-Research-SKILLs
Guides reinforcement-learning post-training of LLMs with slime, which pairs Megatron-LM training with SGLang rollouts, including GRPO runs on GLM, Qwen3 and Llama 3 models.
runzhliu/deepseek-harness-docker
Deploy, configure, verify, upgrade, and troubleshoot DeepSeek Harness with the community Docker, Docker Compose, rootless Podman, and Helm runtime, including the built-in Chromium/noVNC browser…
FreakStudioCN/mpy-hardware-extension
用本地前端插件连云端后端做端到端测试 / test the local VS Code extension against the deployed cloud backend.
majiayu000/spellbook
在已经过独立验证的 CLIProxyAPI upstream 之上部署 NewAPI 计费层,把 Codex/Claude/Gemini/Qwen 等订阅账号包装成可计费的 OpenAI 兼容 API。本 Skill 不负责新建裸 CLIProxyAPI;负责 NewAPI Docker 部署、容器到宿主桥接、模型计费倍率、参数化额度修正、多账号 OAuth…
amd/skills
Serves AI models on AMD Instinct GPU hardware using vLLM. An agent skill from amd/skills.
weave-os/router
Local test harness for the Weave router: a docker compose stack plus codex exec runs that confirm how Codex requests are routed, translated and marked.
weave-os/router
Works through every review comment on a pull request in one pass, fixes what it can, escalates human decisions and keeps CI churn to a single push.
weave-os/router
Correlates a Claude Code session's local transcript with a model router's production cloud logs to explain why a specific response rendered the way it did.
weave-os/router
Installs a language server such as gopls, typescript-language-server, pyright or rust-analyzer and, with explicit confirmation, its underlying toolchain so the lsp tool can use it.
weave-os/router
Correlates a Codex CLI session's local transcript with a model router's production logs to explain why a reply rendered the way it did.
weave-os/router
Shows when to answer a code question through a real language server instead of grep, covering definitions, references, hover, outlines, and errors.
Categories
Stands up the Weave model router in Docker Compose and drives it with claude -p against a real or mocked upstream to reproduce and verify routing and streaming behavior. 1, DeepSeek or Qwen. Two upstream modes are supported, a real provider API that needs working credentials and credits, or a deterministic mock upstream emitting an exact SSE shape with no credits required.
Weave Router Local Testing fits situations like: verifying a router fix end to end against a specific upstream model; reproducing a production routing bug locally; confirming a model's streaming behavior such as tool-call suppression or loop breaks; testing a /force-model route without touching global Claude Code config.
Run `npx skills add weave-os/router --skill test-claude-locally -a claude-code`. Or copy the skill folder (.agents/skills/test-claude-locally in weave-os/router) into .claude/skills/test-claude-locally in your project. Claude Code loads it when a task matches its description.
Run `npx skills add weave-os/router --skill test-claude-locally -a codex`. Or copy the skill folder (.agents/skills/test-claude-locally in weave-os/router) into .agents/skills/test-claude-locally in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add weave-os/router --skill test-claude-locally -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/test-claude-locally, .gemini/skills/test-claude-locally, .github/skills/test-claude-locally and .opencode/skills/test-claude-locally in your project.
Going by SKILL.md and its folder, Weave Router Local Testing needs Python for the scripts in its folder, the command-line tools its instructions call (docker, claude, python3, curl, make and psql) and credentials named FIREWORKS_API_KEY and ANTHROPIC_API_KEY. Our summary lists: Docker Compose; The claude CLI; A real upstream API key and credits, or the mock upstream script.
SKILL.md contains no URLs. Its commands use docker, curl and git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
Weave Router Local Testing is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.1k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Weave Router Local Testing: LLM Council on Fireworks AI (dair-ai/dair-academy-plugins, 614 stars), slime RL Post-Training (Orchestra-Research/AI-Research-SKILLs, 13k stars), Deepseek Harness Docker (runzhliu/deepseek-harness-docker, 110 stars) and Cloud Test (FreakStudioCN/mpy-hardware-extension, 154 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
weave-os (a GitHub organization) maintains it in weave-os/router, which has 5,578 GitHub stars. The repository holds 19 skills in this directory. The repository was last updated on October 10, 2026.
Source: weave-os/router on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.