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 agentmods add instructions/podden/openwebui_anthropic_api_manifold_pipe/agents-mdgit clone --depth 1 https://github.com/Podden/openwebui_anthropic_api_manifold_pipeWrote 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/instructions/podden/openwebui_anthropic_api_manifold_pipe/agents-md)<a href="https://agentmods.dev/instructions/podden/openwebui_anthropic_api_manifold_pipe/agents-md"><img src="https://agentmods.dev/badge/instructions/podden/openwebui_anthropic_api_manifold_pipe/agents-md.svg" alt="Measured on agentmods" 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 | $0.02221 | $0.02221 |
| Opus 5 | $0.01111 | $0.01111 |
| Sonnet 5 | $0.00444 | $0.00444 |
| Haiku 4.5 | $0.00222 | $0.00222 |
Grade A, and why
openwebui_anthropic_api_manifold_pipe AGENTS.md scanned grade A with 0 findings 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 6d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Anthropic API Manifold Pipe — agent notes
Scope
This repo holds one OpenWebUI manifold pipe for the Anthropic Messages API plus its companion/toggle filters. Tasks usually touch model discovery, request payload assembly, the tool loop, content-block streaming, prompt caching, Files API / code execution / skills, or the toggle filters.
Golden rule
anthropic_pipe.py and anthropic_pipe.min.py are generated (both committed). Edit src/anthropic_pipe/**, then rebuild. Hand-edits to the generated sections are lost on the next build.
Build
| Need | Command (from repo root) |
|---|---|
| Compile the single-file artifact | python helpers/build_anthropic_pipe.py |
| Minify it standalone (the build already does this) | python helpers/minify_pipe.py anthropic_pipe.py -o anthropic_pipe.min.py --check |
| Pull the current artifact back into the template | python helpers/build_anthropic_pipe.py --refresh-template |
| Check model-list caching and invalidation | python helpers/test_model_cache.py (fakes the Anthropic client, no network) |
| Check API key encryption at rest | python helpers/test_valve_encryption.py (runs against the built artifact) |
--refresh-template overwrites src/anthropic_pipe/pipe_template.py with the current artifact minus its generated sections. Use it only when the template drifted, never as a normal step.
The build works by marker pairs (# BEGIN/END GENERATED SECTION: anthropic_pipe.<section>) inside the template. Marker order in the template pins section order in the artifact — a module must appear above any code that annotates against it, because the build strips from __future__ import annotations.
File map
| Path | Role |
|---|---|
src/anthropic_pipe/pipe_template.py |
Template for the pipe module: docstring/frontmatter (title, version, requirements), class Pipe, valves, generated-section markers. Version bumps and changelog live here. |
src/anthropic_pipe/request/payload.py |
Payload assembly source of truth: messages, tools, betas, thinking, Files API, context management, cache-control inputs. Compiled into a module-level create_request_payload(). |
src/anthropic_pipe/request/{cache_control,messages,tools,files,rag}.py |
Pipe method groups for the request half of a turn. |
src/anthropic_pipe/response/*.py |
Response half: one module per Anthropic content-block family (text_block, thinking_block, client_tool, server_tool, code_execution_results, web_tool_results, internal_tool_results, compaction_block, status_events) plus state, handlers, registry, runtime, formatting. |
src/anthropic_pipe/shared/{models,tasks}.py |
Model discovery/capabilities and OpenWebUI task handling (title, tags, follow-ups, memory review). |
src/anthropic_pipe/pipe_orchestrator.py |
Pipe.pipe() request orchestration and the high-level tool loop. |
helpers/build_anthropic_pipe.py |
Compiles src/ into anthropic_pipe.py. |
helpers/minify_pipe.py |
Strips comments/docstrings (keeps the module docstring OpenWebUI parses) for the upload artifact. |
anthropic_pipe.py |
Generated single-file OpenWebUI artifact (committed; this is what users install). |
anthropic_pipe.min.py |
Generated minified artifact. Committed too -- it is offered as an install source, so it must never lag behind anthropic_pipe.py. The build writes both. |
anthropic_manifold_companion_filter.py |
Routes OpenWebUI's built-in web_search / code_interpreter buttons to Anthropic-native tools. |
anthropic_pipe_{thinking,web_search,code_execution,files}_toggle.py |
One-shot toggle filters. |
README.md |
User-facing docs: install, valves, changelog. Keep the changelog in sync with the template docstring. |
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.
- 6d ago First seen · 99 lines · 2,221 tokens per session scan A 5acbe8eb7a96
openwebui_anthropic_api_manifold_pipe AGENTS.md is an instructions file published in the GitHub repository Podden/openwebui_anthropic_api_manifold_pipe (52 stars, last pushed 9d ago), licensed MIT. It adds 2,221 tokens to every session, about $0.0111 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).