tesseract-agent 0.11.26 → 0.11.27
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +34 -3
- package/dist-bundle/cli.js +48846 -40649
- package/dist-bundle/mesh-core/package.json +1 -1
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -36,7 +36,7 @@ The service auto-starts at login/boot and restarts on crash: **launchd** LaunchA
|
|
|
36
36
|
on macOS, **systemd `--user`** unit on Linux (with `loginctl enable-linger` so it
|
|
37
37
|
survives logout), **Task Scheduler** ONLOGON task on Windows. Run `install` as your
|
|
38
38
|
normal user — not `sudo` (a per-user service needs your own session, and finds your
|
|
39
|
-
credentials
|
|
39
|
+
credentials and your frameworks' own sign-ins in your home directory).
|
|
40
40
|
|
|
41
41
|
## Multiple agents under one user
|
|
42
42
|
|
|
@@ -68,5 +68,36 @@ docker run --rm \
|
|
|
68
68
|
Or pre-provision credentials and mount them (persist `~/.tesseract` or set
|
|
69
69
|
`CCB_DAEMON_DIR` to a mounted volume), then just `tesseract-agent run`.
|
|
70
70
|
|
|
71
|
-
|
|
72
|
-
|
|
71
|
+
## Frameworks and models
|
|
72
|
+
|
|
73
|
+
The agent isn't tied to one agent framework or one model vendor. You pick the
|
|
74
|
+
**framework** that runs the agent, and the **provider** whose models it talks to.
|
|
75
|
+
|
|
76
|
+
**Framework.** `run --engine <id>` picks it: `claude-code` (the default),
|
|
77
|
+
`openai-codex`, `opencode`, `gemini-cli`, `qwen-code`, `kimi`, `cursor`, `hermes`,
|
|
78
|
+
`openclaw` and the rest Tesseract supports. The framework's own CLI must be on the
|
|
79
|
+
machine. `tesseract-agent assist install <framework>` installs it, and `--engine-path`
|
|
80
|
+
points at a copy that isn't on `PATH`.
|
|
81
|
+
|
|
82
|
+
**Provider.** A runner config pairs a framework with a model provider and its
|
|
83
|
+
credential: Anthropic, OpenAI, Google, OpenRouter, DeepSeek, Moonshot, Z.ai,
|
|
84
|
+
MiniMax, xAI, Azure OpenAI, Bedrock, Vertex, a local Ollama or LM Studio, or any
|
|
85
|
+
OpenAI-, Anthropic- or Gemini-compatible endpoint (`--provider custom-openai
|
|
86
|
+
--base-url …`).
|
|
87
|
+
|
|
88
|
+
```bash
|
|
89
|
+
tesseract-agent runner add --engine opencode --provider openrouter \
|
|
90
|
+
--secret '$OPENROUTER_API_KEY' --model <model> --select
|
|
91
|
+
tesseract-agent runner login claude # add a Claude subscription
|
|
92
|
+
tesseract-agent runner list
|
|
93
|
+
```
|
|
94
|
+
|
|
95
|
+
A secret given as `$VAR` is read from the environment when the agent runs, so it is
|
|
96
|
+
never written to disk. Add several configs and the agent rotates between them when
|
|
97
|
+
one hits a rate limit, a quota or an auth error, across providers and frameworks.
|
|
98
|
+
The running agent picks up changes on its next turn.
|
|
99
|
+
|
|
100
|
+
With no runner config, the framework uses whatever sign-in it already has on this
|
|
101
|
+
machine: a `claude` login or `ANTHROPIC_API_KEY` for Claude Code, a `codex` login
|
|
102
|
+
or `OPENAI_API_KEY` for Codex, and so on. `--engine-env KEY=VAL` passes extra
|
|
103
|
+
credentials to the framework.
|