Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add air-gapped/skills --skill vllm-tool-parsersgit clone --depth 1 https://github.com/air-gapped/skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/air-gapped/skills/vllm-tool-parsers)<a href="https://agentmods.dev/skills/air-gapped/skills/vllm-tool-parsers"><img src="https://agentmods.dev/badge/skills/air-gapped/skills/vllm-tool-parsers/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/air-gapped/skills/vllm-tool-parsers"><img src="https://agentmods.dev/badge/skills/air-gapped/skills/vllm-tool-parsers.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00133 | $0.05598 |
| Opus 5 | $0.00067 | $0.02799 |
| Sonnet 5 | $0.00027 | $0.01120 |
| Haiku 4.5 | $0.00013 | $0.00560 |
Grade A, and why
vllm-tool-parsers scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sS $VLLM/v1/completions -H 'content-type: application/json' \ How it starts
The opening of the file, as written. The whole thing — 261 lines — stays where its author put it; the contents beside it link to each section on GitHub.
vLLM Tool Parsers — Navigation Map
This skill points to the right source file, template, or GH issue. The source code is authoritative — read it. Do not paraphrase from this skill when the actual file is available.
Where things live
Assume a local vllm-project/vllm checkout is accessible. Every reference below is relative to that repo root.
| Target | Read |
|---|---|
| All tool parsers | vllm/tool_parsers/ (one file per parser) |
Parser base class + ToolParserManager |
vllm/tool_parsers/abstract_tool_parser.py |
Shared helpers (partial_json_loads, find_common_prefix, make_valid_python, partial_tag_overlap, compute_tool_delta, handle_single_tool) |
vllm/tool_parsers/utils.py |
| Built-in parser registry | vllm/tool_parsers/__init__.py — _TOOL_PARSERS_TO_REGISTER maps CLI name → module → class |
| Unified parser engine (new) | vllm/parser/ — one class per model (qwen3.py, gemma4.py, deepseek_v4.py, deepseek_v32.py, seed_oss.py, …), abstract_parser.py, and engine/ (parser_engine.py, streaming_parser_engine.py, incremental_lexer.py, token_id_scanner.py) |
| Adapter construction | vllm/parser/engine/registered_adapters.py — make_adapters(XParser) returns (XParserReasoningAdapter, XParserToolAdapter); the tool side is then subclassed in vllm/tool_parsers/*_engine_tool_parser.py to attach structural_tag_model |
| CLI flag definitions | vllm/entrypoints/openai/cli_args.py — grep tool_call_parser, enable_auto_tool_choice, tool_parser_plugin |
| Non-streaming serving invocation | vllm/entrypoints/openai/chat_completion/serving.py — grep extract_tool_calls |
| Streaming serving loop + tail flush | same file — grep extract_tool_calls_streaming, prev_tool_call_arr |
| Plugin import wiring | vllm/entrypoints/openai/api_server.py — grep import_tool_parser |
| Responses API tool handling | vllm/entrypoints/openai/responses/serving.py + vllm/entrypoints/openai/parser/responses_parser.py |
| Per-parser Jinja chat templates | examples/tool_chat_template_<family>.jinja |
| Per-parser tests (executable spec) | tests/tool_parsers/test_<name>_tool_parser.py + tests/tool_parsers/common_tests.py |
| User-facing docs | docs/features/tool_calling.md |
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 261 lines · 133 tokens per session scan A fc915509165f
vllm-tool-parsers is a skill published in the GitHub repository air-gapped/skills (5 stars, last pushed 10d ago), licensed MIT. It adds 133 tokens to every session and 5,598 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
tg-alerts
Use when adding Telegram error notifications to any project. Guides through bot creation with @BotFather, chat/channel/forum-topic ID discovery, alert service implementation with deduplication and graceful failure, and framework-specific integration for Python (async/sync) and Node.js.
zasilkovna
Integrate and manage Zásilkovna (Packeta) shipping for Czech and Slovak logistics. Create shipments, generate labels, track packages, manage pickup points, and configure webhooks using the Zásilkovna API. Covers both the REST API and SOAP API. Includes ready-to-run CLI scripts for all API operations. Context: User…
fastapi-patterns
FastAPI patterns for async APIs, dependency injection, Pydantic request and response models, OpenAPI docs, tests, security, and production readiness.
stripe-projects
Provision SaaS services + sync creds via Stripe Projects.
gemini-api-agent-platform
Guides the usage of the Gemini API on Agent Platform with the Google Gen AI SDK for enterprise AI applications. Covers SDK usage (Python, JS/TS, Go, Java, C#), capabilities like Live API, tools, multimedia generation, caching, and batch prediction.
cqrs-implementation
Implement Command Query Responsibility Segregation for scalable architectures. Use when separating read and write models, optimizing query performance, or building event-sourced systems.