- userland/bundle.nu builds the userland and assembles target/userland/bin
(bin names, tools.txt aliases, brush as sh and bash).
- run and sh take --userland (default SAUSAGE_USERLAND); /bin is read-only.
- The console's bash tool runs through the process table with the
conversation id; large output spills to the world's /tmp.
- sausage run starts the services supervisor.
P1 exit session (openai/gpt-6.1-sol, fresh world): write, read, edit, git
init/add/commit/diff/log, grep -rn, rg, find and a sed|sort|uniq|wc pipeline.
rg (ripgrep 15.2), grep (ours, on ripgrep's regex crates), find and xargs
(ours), sed (uutils 0.3 wrapper), diff and cmp (uutils diffutils 0.6 with -r),
fd 10.5, jaq 2.3, env and timeout. userland/spawn is a small Command-like
client of the process interface. search-test.nu: 42 cases match native tools.
- Core modules run on the host-side preview1 over the shared VFS (D41).
- wasi:http default handler in the linker; anonymous HTTPS git clone works.
Sockets stay denied until P2.
- Each process carries the conversation (from Exec, inherited by children,
shown in /proc/processes) and a host-assigned SAUSAGE_PID.
- Spawn requests choose pipe, inherit or null per stream.
- /config/services.yaml supervisor: live edits, restarts, crash limit 5/10 s,
graceful stop (stdin EOF, then kill after 5 s) (D44).
- Compile caches keyed by content; PATH=/usr/local/bin:/bin.
world.rs also carries T13's optional read-only /bin mount (D43).
T8: sausage:process WIT (spawn, wait, kill), one store per process over the
shared root, /proc/processes, /run, pipeline and cancellation groups, Exec.
brush and 78 uutils commands; the shared cwd-init object sets the cwd from PWD
and keeps numeric exit codes. The corpus matches native bash.
T9: V8 (deno_core) runs pi-durable, pi-ai and chord; SQLite storage in
<dir>/host/harness.sqlite (WAL, synchronous=FULL); file tools on the VFS, bash
through Exec; resumes after SIGKILL.
Preview1 state lives in the host, so Go modules (esbuild) work. esbuild bundles
byte-identical to native; uutils and QuickJS-ng file tests pass. The failed
guest-side adapter attempt (T7) moves to spikes/s7-preview1-adapter/.
Local workflow, smart-HTTP clone/fetch/push and public HTTPS clone/fetch match
native git in stdout and exit codes. Only WASI imports (wasi:http, exit-with-code).
The workspace is now host/ and vfs/. v2 crates leave the workspace as reference.
host/src/exec.rs is the contract between the process table and the harness.
Write new journal entries under logs/core and merge read-only legacy
history without moving SQLite/WAL files. Atomically spool failed batches,
retaining unspoolable entries for event-driven retries. Journal full trap
diagnostics before UTF-8-safe status display. Port session overflow archives
and followable line/byte excerpts to the still-shipped embedded agent.
Document retention/recovery and cover the contracts with regressions.
Include one rustfmt-only normalization in the existing SDK WASI test.
Card: SAU-143
Make legacy core.exit share core.restart's supervisor gate and pending
request. Seed the stable agent prompt and boot comments with the approval
and shutdown contract, and check real init output under a deadline.
Cover both aliases' gate, response grace, coalescing and exit status.
Card: SAU-135
Return success before scheduling the existing shutdown path, coalesce pending
requests, and mark #loop invocations with SAUSAGE_LOOP=1. Keep core.exit
compatibility and document explicit per-call approval for all agents.
Card: SAU-134
- thinking (off, low, medium, high, xhigh, max) is an enum: serde rejects
unknown levels in the config, and the schema of agent.chat lists them.
reasoning_effort, a second name for the same setting, is gone.
- agent.chat takes an optional thinking level for one turn.
- off really switches thinking off: OpenAI gets the effort none, and
Anthropic gets disabled thinking (newer Claude models think by
default). A model that cannot switch thinking off gets low.
- Unset means the model's default. Anthropic thinking blocks now return
unless thinking is off, because newer models think by default.
- The template sets thinking: medium and shows it in each example, with
current model names.
- Checked against the proxy with gpt-6-luna, glm-5.3-flash, and
claude-haiku-5-5.
- /config/agent.yaml and /config/boot.yaml replace the JSON files, read
with serde_yaml_ng (also in the agent, inside WASM).
- sausage init asks nothing and reads no SAUSAGE_* variables. It writes
commented templates with an example for each API, and prints the path
to edit. Until the agent is configured, agent.chat names the missing
field. allow_no_token: true permits a local model without a key.
- The agent speaks openai-chat-completions, the format of OpenRouter,
Ollama, and most providers. base_url follows each SDK: with /v1 for
the OpenAI shapes, without it for Anthropic.
- Checked against the proxy with glm-5.3-flash: the agent wrote, built,
and activated a counter plugin in 55 s.
mold applies to the host target only, in the dev shell, the crane builds,
and the packaged binary (proc macros and build scripts of plugins). A
rebuild of the host test binary after a change takes 0.9 s (15 s with
GNU ld). Measured LTO: no gain worth 4 more minutes per release build.
The console reads a line, runs it to the end, and prints the result,
with and without a terminal. Notifications print while a command runs.
Ctrl-C during a command stops only the wait. 314 lines (was 480).
All host integration tests are modules of cli/tests/host/, so they run
in one parallel pool, and the probes build once. Warm, the 184 host
tests take 24 s (was about 45 s as seven binaries in sequence), and
cargo test --workspace takes 45 s.
128 fast activate/revert cycles are not a realistic load. restarts_service
covers the restart of a service after an upgrade. nix flake check runs
the normal tests only.
The timing rules are in two tests (upgrade_and_discovery_deadlines,
stop_deadlines). All other tests use a grace period of 0 or core.stop,
poll with deadlines instead of sleeping, and use the persistent cache in
CARGO_TARGET_TMPDIR. The 128-cycle singleton test is an ignored stress
test that nix flake check runs; the normal run has a 4-cycle sample.
Warm, the activation suite takes 12 s (was 47 s), and cargo test
--workspace takes 65 s.
surma
Sausage