Reference
Lookup tables for venom as a whole, covering environment variables, ports, error codes, x-venom headers, limits, a glossary, make targets and the design records.
This section holds the tables you look things up in. The guides in the other sections explain how to use a feature; these pages list every name, default and code in one place.
| If you need | Read |
|---|---|
| What an environment variable does and its default | Environment variables |
| Which port a service listens on, and how the pieces connect | Architecture, ports and contracts |
What an error code means and what to do about it |
Error codes |
What an x-venom-* header does, on a request or a response |
x-venom headers |
| Body size caps, timeouts and other built-in limits | Limits and defaults |
| What a word like gear, route profile or daemon means | Glossary |
The generated API reference lists every HTTP operation with its fields and a curl line: the orchestrator (/docs/api/orchestrator), venom serve (/docs/api/harness-server) and Venom Cloud (/docs/api/cloud). The pages here explain behaviour the generated pages do not show, such as which header changes routing or which error is safe to retry.
Conventions on these pages
- Names in
codeare exact: an environment variable, a header, a config key or an error code, spelled as the software reads it. - "Unset" means the variable is absent or empty. Venom treats an empty value as unset unless a row says otherwise.
- Defaults are the values in the code at the time of writing. When a guide and a reference table disagree, the table is the one checked against the source.
For agents
Every page is also plain markdown at the same URL plus .md, for example /docs/reference/errors.md. The error and header tables are written so an agent can match a response it got against them line by line.
In this section
- Environment variables: Every environment variable the venom CLI, the harness and the dev stack read, with what it does and its default.
- Architecture, ports and contracts: How venom's pieces fit together, which port each one listens on, and which file is the contract between them.
- Error codes: The error shapes and codes of the orchestrator, Venom Cloud, venom serve and the CLI, with what each means and whether a retry can help.
- x-venom headers: Every x-venom-* HTTP header venom sends or answers with, on the orchestrator, Venom Cloud and venom serve, and what each one changes.
- Limits and defaults: The built-in size caps, timeouts, queues and counts of the orchestrator, venom serve, Venom Cloud and the CLI, and where to change each one.
- Glossary: Short definitions of the words venom's docs, config and APIs use, from alias and gear to daemon, relay and lease.
- Design decisions (ADRs): One line per architecture decision record, saying what venom decided and why it matters to you.
- make targets: What each make target in a venom checkout does, for running the dev stack, building and testing.
- Audit and compliance export: Export one period's evidence of who approved what with venom audit export, and check it offline with venom audit verify-export.
- Single sign-on and SCIM for Venom Cloud: Connect a Venom Cloud organisation to Okta, Microsoft Entra ID or Google Workspace over OIDC or SAML, and let SCIM keep the member list in step.
- Reasoning-effort auto-tuning: Let the orchestrator pick a reasoning effort per request with reasoning_effort auto in a route profile, and what it will never override.
- Evidence bundles: Print the proof a change carries with venom evidence, as JSON or a pull request comment, and build it from code with @venom/harness/evidence.
- Headless output with --json: Drive venom commands from CI, editors and other agents with --json, which prints one JSON document on stdout per command.
- Intent ledger: Record what is wanted as an intent with goals, constraints and acceptance checks, and let agents claim it, attach evidence and close it with venom intent.
- Knowledge Q&A: Ask how your team works with venom ask, the console's Ask page or Slack, and get an answer cited from org memory, the team library, session summaries and your repo's docs.
- Onboarding paths: Build guided first tasks for new members on Venom Cloud, and walk them with venom onboard.
- Public mesh hardening: How the public mesh checks that a daemon run by a stranger serves the registered model at the promised quality, what was measured, and what is still weak.
- Stacked pull requests with venom stack: Split a large branch into a chain of small pull requests, submit, restack and land them with venom stack.
- Browser IDE: Review and steer an agent session from a browser with the small IDE venom serve hosts at /v1/ide, and what it can and cannot do.
- How these docs are built: Build and check the public docs from a venom checkout, serve them from your own Venom Cloud, and read them as markdown, llms.txt or search.json.
- The offline API and SDK docs site: Build a static, dependency-free docs site from venom's own contracts with make docs-site, for air-gapped networks, and check it with make docs-site-check.
- Realtime voice agents (design): The design for phone and browser voice agents on the harness, such as a receptionist or a collections agent, and where each part will run. Not built yet.
- Research behind the design: The cited research venom's design decisions rest on, how references are keyed, and the study of opencode that fed the backlog.