Files
deepseek-harness/packages/examples/agent-spine-demo/README.md
T
Tianyi Cui 6143537d4e Merge branch 'codex/simp-agent-entry-state' into codex/simp-unify-agent-session-id
# Conflicts:
#	docs/cordis-catalog/events.md
#	docs/cordis-catalog/services.md
#	docs/core-data-structures/bash.md
#	docs/core-data-structures/core.md
#	docs/event-producer-consumer.md
#	docs/module-graph.md
#	packages/bash/bash/src/types.ts
#	packages/bash/tool-bash/README.md
#	packages/bash/tool-bash/src/index.ts
#	packages/bash/tool-bash/tests/tools.spec.ts
#	packages/examples/agent-spine-demo/README.md
#	packages/examples/agent-spine-demo/tests/agent-core.spec.ts
#	packages/subagent/tool-subagent/tests/tool-subagent.spec.ts
#	packages/util/brand/README.md
#	packages/util/brand/src/index.ts
2026-07-15 23:29:47 +08:00

63 lines
5.2 KiB
Markdown

# @deepseek-ai/dsh-agent-spine-demo
The **default executor-less, UI-less agent spine** as ONE Cordis bundle plugin. It loads the fixed set of services every harness agent needs, including the local skill provider, and forwards the loop's `agents` list as its own config — so an app package composes a working agent by adding only a front door and the swappable backends.
Read this package for the whole plugin tree and its composition order.
## The tree it loads
`apply(ctx, config)` mounts each of these as a child of the bundle fiber:
```
@cordisjs/plugin-timer timer service (writes nothing to stdout)
@deepseek-ai/dsh-llm abstract LLM service + content-block vocabulary
@deepseek-ai/dsh-session event-sourced session log + store
@deepseek-ai/dsh-system-prompt prompt-section + tool-schema assembly
@deepseek-ai/dsh-tools registry + guarded pre/around/post/final-result pipeline
@deepseek-ai/dsh-skill skill provider registry
@deepseek-ai/dsh-skill-local local filesystem skill provider
@deepseek-ai/dsh-agent agent registry + agent/* event vocabulary
@deepseek-ai/dsh-tasks generic background-task registry
@deepseek-ai/dsh-invariants dev-mode event-contract assertions
@deepseek-ai/dsh-tool-bash the model-facing bash schema
@deepseek-ai/dsh-tool-skill session-prefix skill catalog + model-facing loader schema
@deepseek-ai/dsh-tool-tasks task_output/task_list/task_kill schemas + completion notices
@deepseek-ai/dsh-agent-loop THE concrete loop (gets the forwarded `agents`)
(dsh-system-prompt gets the forwarded `persona`)
```
## What it deliberately leaves OUTSIDE the bundle
The spine is everything COMMON to every front door. The swappable and front-door-coupled pieces stay out, picked by whatever loads the bundle:
- **the LLM adapter** — the bundle ships the abstract `llm` service; the leaf registers a concrete adapter on `ctx.llm` (`llm-deepseek`, `llm-pi-ai`, `llm-replay`).
- **the bash executor** — the bundle ships `tool-bash` (the consumer schema); the leaf provides `ctx.bash` (`bash-local` or a sandboxed impl).
- **non-local skill providers** — the bundle ships the skill registry, the local filesystem provider, and the `skill` tool; deployments can add other providers such as embedded or remote catalogs as siblings.
- **presentation + per-app infra** — the stdio UI / ACP bridge, a console logger, `hmr`. These form the coupled "front-door cluster" that the app packages ([`dsh-stdio-demo`](../../examples/stdio-demo/README.md), [`dsh-acp-demo`](../../examples/acp-demo/README.md)) bake in. `timer` is in the spine (common to both, stdout-silent); a console logger is NOT (it writes to stdout, which the ACP bridge reserves for JSON-RPC).
This is the [interface/implementation/consumer seam](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md) raised to the composition level: the bundle owns the shared spine, the leaf owns the backends, the app package owns the front door.
## Config
```ts
import type { Config } from '@deepseek-ai/dsh-agent-spine-demo'
// { agents?, persona?, toolOrder?, tools?, skills?, toolBash?, toolTasks? }
// The schema intersects the owner schemas,
// so validation and defaulting can never drift from the owners.
```
The bundle FORWARDS each field to the child that owns it: `agents` to `agent-loop` (default `[]`), so each app supplies its own pre-created agents — a stdio app pre-creates one under the `main` config label; the ACP app pre-creates none (it creates agents on demand at `session/new`) — `persona` and `toolOrder` to `dsh-system-prompt`; `tools` to the tool registry for its presentation mode; `skills.registry`, `skills.local`, and `skills.tool` to the skill registry, local provider, and model-facing consumer; and `toolBash`/`toolTasks` to the two model-facing tool plugins the bundle owns. `toolBash.enableRunInBackground` controls only the bash producer, while `toolTasks` controls generic `task_output` wait bounds; independently loaded producers keep their own config. Forwarding is exactly why the owners can live in the shared spine even though the apps disagree on what to configure.
## Why a code bundle, not a shared YAML include
A YAML include can deduplicate config but cannot own a bin or provide front-door defaults. App packages make stdout-safe ACP wiring the default, though a leaf can still add an unsafe logger. Bundle children register services in the root isolate-keyed store, so injected leaf siblings see them without load-order coupling.
## Model Experience
Indirectly, through `dsh-system-prompt`, `dsh-tool-skill`, `dsh-tool-bash`, and `dsh-tools`, which this bundle mounts without adding model-bound wrapper content.
## Known Limitations and Deferred Work
- **The spine set is fixed in code** — `apply()` mounts every child unconditionally (including `tool-bash`); no config excludes or replaces one, so swapping the loop or dropping a spine member means composing a different bundle.
- **`dsh-invariants` mounts unconditionally** — this bundle has no toggle, so every composition using it pays the dev-mode relational assertions; Session's always-on validation and freezing are separate.