A run that repeats itself, or keeps getting the same correction, ends in
code as escalated instead of running to its turn cap (L2, M7.8), and the
page flags such runs (X6). While the call record is on, each system
prompt says so.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
nisse-ui (X6) shows the operator each agent's work and, per item, a
walkthrough down to every model request and response, for studying how
context, compaction and memory work. It reads each agent's state store,
call record and memory folder through read-only mounts, so agent code is
unchanged and each agent stays the only writer of its state; it is
released with the agents and tested against stores their own code wrote.
The call record (L7) is new in nisse-core: every request as sent, the raw
response with its reasoning, timing, tokens, prompt reuse and each
message's provenance. It is diagnostic only, never read by the agent, and
deleted at start, so it never migrates.
Commands from the page, including a hold on a task or a whole agent, are
an open question (Part Q). The shared database first proposed in #109 is
dropped. The plan gets M6.6.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
A `review` target in the Dockerfile: its own user (nisse-review, uid
10002), git for its clones, state and clones in /var/lib/nisse-review,
configuration at /etc/nisse-review, and only its own agent installed.
release.yml publishes gitea.dongar.net/dev/nisse-review beside nisse-dev,
at the same version. review-bringup.sh builds and runs the image, as
bringup.sh runs nisse-dev's.
The test stage adds an agent package its base lacks, without its
dependencies, since the suite's own code imports every agent's. Without
that, the tests over nisse-dev's image no longer import since
test_review_live arrived. tests/deploy/test_images.py checks the image and
that the test stage leaves what it installed as it was.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
nisse-review's review of #169: a reply quoting a top-level point, or one
Gitea refused on its line, arrives through list_comments, so answering it
would start the thank-you loop again. Also records what the live Gitea
showed: resolving is per comment, and resolving again is a 204.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
The reviewer is read-only (R3), so Gitea refuses it resolving a thread.
Now each reply says whether it settles its thread, and a settled reply
carries <!-- nisse-review: settled -->. nisse-dev, as the author, resolves
every comment on that thread's line when it watches the PR, and doesn't
take the settled reply as a point to answer. scripts/resolve_settled.py
does the same for PRs made by hand.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
On #166 the reviewer called post_review with two points unanswered, was
refused, answered them, then only skipped already-answered points until the
40-turn cap: the review went out as partial, its summary lost. A refused
ending call is now held, and the driver ends the run with it as soon as the
last point is settled (a new due hook in nisse-core's driver; nisse-dev
doesn't use it). A held call that is wrong for another reason by then
(sign_off over drafted comments) is left to the model. With nothing open,
reply and skip say which tool ends the run.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
The operator chose single-space mode: the vault is mounted at /data, with
memory/ read-only over it, and there are no accounts or setup wizard;
Authelia in the operator's Caddy is the login. SB_NAME forces the mode, so
an empty vault can't boot the wizard. SilverBullet's own data folder is
gone. check now fails if the wizard could be open. The README says how to
make memory/ writable for a while and back.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Each agent ships as its own image, since that's how it runs on the agent
host. nisse-review's image becomes M7.6, ahead of the agent host; its
deployment moves to M7.7. nisse-ops' image comes with its first slice.
Also fixes two stale references to M7.6.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
deploy/agent-host sets up the vault on nisse-host (infra/tofu#77) and
SilverBullet over it: the operator writes the vault, the agents' memory/
is read-only to it, and only the operator's Caddy reaches it, which
serves vault.dongar.net to the LAN. scripts/agent-host.sh installs and
checks it from here.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Gitea dismisses an approval made stale by a push in a job it queues on the
push (services/pull AddTestPullRequestTask, v28.1.0), and its merge check
counts approvals not yet dismissed. Measured here, the approval stands for
0.15-0.23 s after the push; the test's merge came a couple of API calls
later, so it usually lost the race, and once won it. The test now waits for
the dismissal, with a timeout, before expecting Gitea's refusal, and checks
the agent's own rule at once: the PR's head is read from the branch, so the
approval is for an older commit straight away.
Also notes in the Development Plan why the helper loop doesn't compact.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Before each driver call, over compact_at (0.6) of the prompt budget, the
same model is asked for notes over the very request it was about to get,
plus one message, so the server's cache covers all but the notes. The notes,
with code's list of the calls they replace, become a compaction event that
renders in place of its range, keeping word for word what started the work
and the last few turns. Rare, large steps keep the prompt's start unchanged
between them. A later compaction folds in the earlier; one from an earlier
run's view is hidden.
Failed notes are logged and fitting takes over. Fitting now applies
compactions before it measures, records what it cuts (Trim) and logs each
cut as context.trimmed. Wired into both agents' drivers and the benchmark;
nisse-dev's helper loop is a follow-up.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Archives were downloaded and unpacked afresh for every run, which doesn't
scale to very large repositories. Each repository is now cloned once, with
its whole history, under scratch/repos/<owner>/<name>/ (head/ the clone,
base/ a worktree when read). A run fetches every branch and the PR's
refs/pull/N/head, checks out the head exactly, and fails if that head is
gone. Only the latest repository is kept for now.
Git runs walled off: none of the host's git configuration, links checked
out as plain files, no hooks, submodules or LFS, only the Gitea URL's
protocol, and the token as an http.extraHeader for that URL in git's
environment only. A stop ends a git command; a cut-short first clone is
removed. The reader never shows .git. GiteaClient.archive is gone.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
SIGTERM used to stop only new rounds: a review in progress ran on until
systemd killed it 90s later. Now the scheduler's stop sets the driver's
cancel flag, so the run ends at its next turn, and a model call in flight
is broken off (Interruptible, raising Stopped from the signal handler, only
ever inside a model call). The item goes back to ready, with nothing
drafted posted and no failure counted; replies already posted carry their
markers and are not repeated.
intake.repo_failed is logged when a repository starts failing or the reason
changes, and intake.repo_ok when it lists again, not every round.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Back on a PR it reviewed: a new head gets a later review of the commits
since (each one's diff, through the API) and of its earlier points; what
was said since gets replies, answered where it was said or skipped with a
reason. It signs off once and stays quiet while still happy. Past
tuning.review.max_rounds it sums up for the operator and goes quiet,
until a push or someone other than the PR's author joins in. An
operator's `stop` quiets it; `review` asks for a fresh review.
Reviews carry numbered markers and replies their point's key, so
recovery posts nothing twice.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
From its failed review of #156: read_file stopped at 20,000 characters
with no way on, so the reviewer hunted each file's end a definition at a
time and ran out of turns; its retry carried every read before it, and a
model call timed out. Both drafts were lost.
- read_file takes a start line and says where the next part starts.
- Each run's view starts at its brief (Transcript.since); the log keeps
the rest.
- A run out of turns with points drafted posts them as a partial review.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
At bring-up, `dev/*` failed: listing an organization's repositories
(/orgs/dev/repos) needs the read:organization token scope, which the
reviewer's token doesn't have and shouldn't need. Patterns are now matched
against /repos/search, which needs only the repository scope and already
lists every repository the account can read. A pattern that matches
nothing is logged (access.no_match). owner_repositories goes.
A Gitea error at startup is now reported like a config error, not a
traceback.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
At bring-up, nisse-review read no repositories: Gitea's /user/repos leaves
out a public repository's Read collaborators, so dev/nisse never showed up.
`repositories` now holds patterns (dev/nisse, dev/*, */*) matched against
every repository the account can read, public ones included (an
organization's or user's list, or the instance's search); unset, it is
/user/repos as before. A configured repository it can't read is left out
with an oplog line.
The read-only wall (R3) is checked every round, not only at startup: an
account granted push or admin while running stops (exit 3) rather than
review on.
nisse-core's Gitea client gains GiteaNotFound for 404s, owner_repositories
and visible_repositories.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
scripts/e2e.sh ran only test_end_to_end_loop.py, and the integration step
excludes e2e-marked tests, so test_review_live.py (M7.3) never ran in CI.
The script now runs every e2e-marked test under tests/integration, and the
review test is a smoke test: about a minute on the local model, and the only
check that the review account, its token and its read-only refusal work.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Anyone's comment on an item's PR but an operator's (and its own) is a
peer's point (D5, L11): kept on the item with an id, in the transcript as
a `peer` message labelled with its author, and never an instruction. The
driver answers each with answer_peer, inline on its line or quoting it,
or skips it with skip_peer and a recorded reason. A run going back to
waiting with points open is reminded once; then the harness lists them
on the PR. A peer's comment wakes the item; a wake by peers alone goes
back to the same wait quietly, with one line in the thread. The
operator's inline comments are now answered inline too.
Also corrects ci.yml's header: a push to main runs no end-to-end tests.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
A new package over nisse-core: it reviews every ready PR in the
repositories its account can read, and posts one COMMENT review.
- scheduler: each round, open PRs per readable repository (or the
configured repositories); an item per PR, ready unless draft or WIP;
closed with the PR. One review at a time, oldest first.
- reader: the repository at head and base from Gitea archives, unpacked
and read in Python (read_file, list_files, search_code), and its issues;
bound to the item's repository; no git, no shell.
- review: starts from the PR, commits, diff, comments, closed issues, CI;
comment drafts a point on any existing line; post_review posts one
COMMENT review at the head, inline where Gitea takes it, else as links in
the summary. ask_expert; vault and memory (memory/nisse-review).
- L8: posting is an intent; recovery asks Gitea for the marked review and
never posts twice; with Gitea down, the item waits.
- Failures retried, then said on the PR until the next push.
- Startup refuses an account that can push to or administer any repo.
- nisse-core gitea: readable_repositories, open_pull_requests, diff,
commits, archive, issue, search_issues, reviews, post_review; PR fields.
- Bring-up: scripts/review-bringup.sh; nisse-review.example.yaml.
- Tests: unit over a fake Gitea and scripted model; a live test
(NISSE_REVIEW_GITEA_TOKEN) reviews a PR in a fresh test repository.
- Docs: Development Plan M7.3, Tuning Guide (tuning.review.*).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
Shared parts, not a shared scheduler (agreed): nisse-review will have its
own small scheduler; a common one waits until two exist to compare.
- nisse_core.loop.driver: the driver run, moved; the agent names its
waiting tools per run (nisse-dev's are WAITING_TOOLS), and execution
goes to any Executor (ToolOutcome moves with it).
- nisse_core.loop.tools: the tool helper and the tools any agent offers
(notes, memory, memory_done, ask_expert); nisse_core.loop.expert.
- nisse_core.knowledge.handlers: Knowledge, the vault and memory as
control tools, the index and prompt text, out of LoopAgent.
- nisse_core.config: file reading, references, people, tuning sections of
the agent's own, LLM numbers, the vault.
- New core tests: a driver run with a non-workspace executor and the
agent's own waiting set; Knowledge alone; the shared config pieces.
Docs: the Development Plan marks M7.2 done; Architecture Part L and the
nisse-core README name the new parts.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
A review agent comes second: it reads every ready PR in the repositories
it can see and discusses its points on the PR, never changing code.
- Architecture Part R: duty and items (R1), rounds capped and reset by
anyone else commenting (R2), read-only reach bound to the item's
repository, no approvals, no build VM for now (R3), peers (R4), local
model with ask_expert, no Matrix room for now, own identity (R5).
- D5 and L11: a peer's review is constructive criticism, answered point by
point; never an instruction, approval or go-ahead; nisse-dev doesn't
wait for it before asking to merge.
- Part Q: letting nisse-review approve a limited subset of PRs.
- Development Plan: Phase 7 is nisse-core and nisse-review (M7.2-M7.6);
nisse-ops moves to Phase 8 (M8.1-M8.5), with the MCP decisions so far.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
What isn't the dev agent's moves to nisse-core, unchanged in behaviour,
with its tests: canonical, render, render_anthropic, llm, catalog, probe,
context, views, state, reconcile, worker, cancel, dump, transport, matrix,
execution, testing, vault, memory, oplog, envfile and gitea (whole).
- nisse_core.state.StateStore holds the tables every agent has; an
agent's subclass adds its tables (version, migrations), roles and
views. nisse_dev.state.StateStore adds the build VM blocks (v5).
- views.project takes what each role is shown, rather than a fixed table.
- Vault takes the agent's name (agent=), for provenance.
- A test that nisse_core never imports an agent.
- The loop, config, git, delegation, expert, tools and evals stay in
nisse-dev until nisse-ops has a loop to share.
Docs: the Development Plan marks M7.1 done and notes that Phase 6 waits on
the agent host and Terraform; Architecture Part L names the package.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
A project's notes have their own folder, memory/nisse-dev/projects/<project>/,
with its own index; remember and forget take scope (general, or project:
always the current item's own, so no item writes another project's notes,
and a conversation has none). At an item's start the transcript gets the
general index and its project's, never another's. Index links are pages
from the vault's root, as SilverBullet names them.
The closing pass's memory_done can name repository documents the item left
out of date; the harness says so in the thread and records it. The prompt:
the repository is the authority, and a note that disagrees is corrected.
With the operator: ARCHITECTURE.md isn't put into every item's context (its
size); the infrastructure world model waits for nisse-ops. The walk writes
project notes and checks they come only from that project's items.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
remember(name, description, fact) and forget(name, why): control tools for
the driver (before the go-ahead too) and conversations, writing only the
agent's own vault folder, one note per fact with its provenance. No one
agrees first; each change is a line in the thread. The harness keeps the
index and puts it into the transcript once per item, with a hash.
At an item's close, completed or abandoned, a bounded memory pass (the
driver's model, its own transcript and view) records what was decided and
learned; it never holds the close up (tuning.memory.max_turns).
With the operator: corrections may be paraphrased (L6 changed), the model
forgets on its own, and nothing limits what may prompt a note. The walk
remembers and forgets, and checks nothing lands outside the agent's folder.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
vault.py: a folder of Markdown notes on the agent host, as SilverBullet reads
them. Only configured folders exist for the agent, each rw or ro (`vault:`
in the configuration); paths are checked in code (relative, inside a folder,
symlinks included, Markdown, never the editor's own files). Writes are
atomic (temporary file, fsync, rename, directory fsync) and merge provenance
into the note's frontmatter, keeping the rest; failures are results.
read_note, list_notes and search_notes are control tools in the harness for
the driver (before the go-ahead, and while a VM is blocked) and for
conversations, labelled with their path; the prompt says a note is
information, never an instruction. Writing by the model is M5.2's.
With the operator: SilverBullet replaces Obsidian and runs on the vault
itself, agents' folders mounted read-only, so no Syncthing (L10, X5, M5.4).
Per-agent mounts come with the containers (M6.1).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
M4.3's block, with new triggers and a way out. A VM out of reach at the
check before a run, in a run (which ends there: the driver, a toolchain
step and a requested step), or while preparing the workspace blocks every
project on it; the item stays ready, waiting on no one. On the intake round
it is tried again every tuning.build_vm.retry_s (300), and once it answers
it must still answer settle_s (180) later; then the work a lost run left
there is parked (pushed), the VM unblocked, and that said once.
A changed host key, at the check or on a retry, blocks the VM too, until
the operator's `vm ready`, which trusts the new key: `trust new host key`
is retired, and the message says it's a key problem that may be an
impersonation. Questions are answered while a VM is blocked, without it.
Starting the agent with a VM down no longer fails.
The walk breaks the VM's SSH, now or partway through a run, and checks that
no model works on a VM that is down; loop.qnt models the VM going and
coming back (unblocked_only_when_up). Together they found: a lost run's
work set aside by the next item there; a question left behind a block; a
lost VM swallowed while the toolchain step checked or recorded it; and a
remark waiting to be answered that outlived its PR's merge. All fixed.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
The driver can request_reimage(reason) when its build VM is broken past the
usual fixes (before the go-ahead too: it asks, it doesn't act). The harness
parks the work in progress (pushed to Gitea), then blocks the VM in the state
store (schema 5, vm_blocks): no item on any project on it is taken up, across
restarts. The operator hears once at top level, with the command to clear it,
and each blocked item's thread gets a line. `vm ready <project>` lifts the
block and trusts the VM's new host key; a reimaged VM is set up again by the
toolchain step, and the item that found it goes first.
loop.qnt gains the block (blocked_vm_runs_nothing) and a scenario; it found
that a reimage lost work in progress, hence parking first. The walk requests
reimages and recovers them, reimaged or not. The unreachable case reuses the
block in #137. Architecture P5 is softer, and D3 describes the block.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01J5Utyfzw7gF7UhjRDEtjTk
The preflight becomes the toolchain step (D3, as #135 describes it): a
bounded run of the item's driver model, in a transcript of its own,
that makes the build VM and the project's files agree.
- Started by the harness when the VM isn't set up for the project
(M4.1's checks): it sets the VM up from the repository and edits
nothing there (no write_file; anything a command changed is reported,
left for the work).
- Requested by the driver with toolchain_step(reason) once work is
agreed: it installs, then brings pins, setup docs and CI
configuration in step; its edits land in the current milestone. The
driver gets the step's report as the tool's result, not its turns.
- D1 made structural: until the go-ahead the fast path refuses what it
knows changes something (and anything dangerous), with a pointer to
the restatement; unknown commands still go to the classifier, so a
bug can be reproduced before it's restated.
- A `toolchain` actor and view; the walk asks for steps now and then.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
D3's preflight and install recording become the toolchain step: a
bounded run of the item's driver model that makes the build VM and the
project's own files (pins, setup docs, CI configuration) agree, with
the repository as the record. The harness starts it when the VM isn't
set up for the project (and then it edits nothing in the repository);
the driver requests it once work is agreed, and its edits land in the
current milestone. Versions stay facts the harness reads itself; a
rebuilt VM's new host key is trusted only on the operator's word.
D1 gains the rule that until the go-ahead the agent only looks:
anything that changes something is refused by the fast path, so what
the work needs goes in the restatement.
The Development Plan's M4.2 follows: the toolchain step, in place of
recording installs.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
claude-haiku-4-5 is being replaced by claude-haiku-5-5 in LiteLLM: the
integration tests' roles and the example config's expert move to it.
Its profile keeps the predecessor's benchmark figures, marked as such
until it's measured. The plan's models-by-role line also gains the
local classifier (M3.2). Results measured on 4.5 stay as they were.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
A model works out and sets up what a project needs on its build VM;
code makes sure that happens, bounds it and checks the result (D3).
- Before each driver run on a project, one round trip checks the build
VM: the harness's own baseline (missing: the item waits, a template
problem), the workspace clone (missing: cloned again and said; the
item resumes from what it pushed) and a per-project marker.
- No marker (a new project, a new or wiped VM): the item's driver model
runs a bounded preflight first, with its own prompt, the execution
tools (screened as ever) and toolchain_ready. The harness runs each
reported version command itself (version-shaped commands only, even
from the marker, which the model's commands could edit) and writes
the marker. Gaps and an unfinished preflight are facts and work goes
on; the thread hears only of installs and gaps. A rebuilt VM starts
from the last toolchain.
- Later runs check the versions again; a change is a fact and a thread
line. The PR body says what the work was tested with.
- A changed host key is its own failure (HostKeyChanged): the item
waits for the operator's `trust new host key`, which drops the pin;
then the VM is set up again.
- A prompt line: install what the work needs on the build VM, say why,
and put lasting needs in the project's own pins.
- tuning.preflight: max_turns, command_timeout_s.
- The property walk gains a VM wipe and ready_before_work (checked
against a mutant that skips preflight); parked_work_safe accepts that
a wipe loses what wasn't pushed (D3).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
A project's trust.merge is operator (the default) or approved (L3, D5,
#65). Set to approved, the harness, never the model, merges a PR in the
PR watch once an approving review from a gitea.operators account is on
its current commit, its checks pass (none at all counts; pending
waits) and it's mergeable; the item then goes on as after any merge.
Each reason to hold off (failed checks, conflicts, a refused merge) is
said once per commit.
Gitea enforces it: protection requires approvals from an operators-only
whitelist and dismisses stale ones. At start, for each approved
project, the agent checks by behaviour that its own unapproved merge is
refused, with a throwaway branch and PR made through the API (never the
workspace); if it isn't refused, the operator merges until a restart.
- gitea: approvals, combined check status, a PR's head commit and
mergeability, API-made branches and their deletion.
- Test and bring-up repositories get the protection.
- Tests: the decision (unit), the hard wall on a real Gitea
(test_trust_live), an e2e test where the operator only approves, and
a walk property, merges_only_approved, checked against a mutant that
ignores which commit was approved.
Closes#65
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
What the fast path calls unknown is judged by an independent model call
on the item's classifier role (L4), before it runs.
- nisse_dev.loop.classifier: it sees the action and the agreed scope
(project, repository, issue number, the restatement the operator said
go to), built in code; never the transcript, the issue's text, tool
output or the driver's reasoning. Its verdict is a tool call: safe
runs; in_scope runs and is reported in the thread; needs_discussion
asks, as a dangerous action does. An error, no verdict or an unknown
one fails closed. Verdicts are cached per item, agreed revision and
action. A helper is refused what needs discussion.
- action_verdict gains source, model and the classifier's exact input;
details shows the source. An allow after the classifier asked is an
override. scripts/classifier-evals.py exports the judgements from a
store, read-only; tests/integration/classifier_cases.yaml is the seed
set, run per model with --capability-models.
- The fast path's delete rule: deleting on the build VM is contained;
only what breaks the host (system and home directories) stays
dangerous, since in tests the build VM is the test machine.
- The classifier role is the local model in the example config.
- The property walk gains a classifier answering at random, junk and
failures included; dangerous_only_when_allowed now covers its
verdicts (checked against a mutant that runs them all).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
Every execution tool call, the driver's and a helper's, gets a verdict
from code before it runs (P3, P4, D6).
- nisse_core.actions, nisse-core's first code: Action and Verdict, and
reading shell commands statically, segment by segment (cd, wrappers,
sh -c). What can't be read that way is unknown; a line is as bad as
its worst command.
- nisse_dev.loop.actions, the rule table: reading, checks and tests are
harmless; workspace edits, local git and installs on the build VM are
contained; pushing, other hosts, uploads, publishing, the forge's own
tools, deleting outside the workspace and wrecking the VM are
dangerous; changing branches by hand is refused (the harness owns
branches) with a pointer to the milestone tools.
- A dangerous call isn't run. The item waits on the operator with a
question rendered from the action, and `allow` lets exactly that
action run once. A helper can't ask, so it is refused and told to
hand back. Unknown actions run, logged, until the classifier (M3.2).
- Every verdict is an action_verdict event, which no model sees, and an
oplog line; `details` shows the counts and every action not simply
run.
- The property walk gains dangerous_only_when_allowed (checked against
a mutant that doesn't use the allow up); the slice e2e checks every
call was judged and none needed an allow.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
nisse-dev now runs only the conversational loop. The gated state machine
it was built alongside is gone:
- Removed control.py, planner.py, executor.py, approvals.py, gates.py
(command_form moves to loop/reader.py), publish.py and poller.py, the
old assembly in agent.py (its shared helpers move to loop/build.py),
the store's state-machine view (get_state, transition), their tests,
the old end-to-end and live planning/milestone/gate tests, and
specs/orchestrator.qnt with its scenarios.
- Roles are driver, helper, expert and classifier; driver is required.
loop:, roles.planner, roles.executor and tuning.executor are refused
with what to use instead; tuning.driver takes the turn cap and failure
limit.
- The benchmark and the capability matrix measure the driver: the
loop's own system prompt and tools, the task's plan as the agreed
restatement. Task plans end with finish_milestone, so earlier runs
show as (old); stored runs' executor tuning reads as the driver's.
- A delegation on the loop records what its helper's alias resolved to
(L9), as the old path's did.
- The dump shows a work item's revision and what it waits for.
- Behaviour the old tests covered and the loop shares stays covered:
driver runs that fail on a model call, stop mid-turn, reset the
failure count; go-aheads ignoring case and punctuation; a request too
big to fit is never sent; the assembly tests, now on build_loop_agent.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
- The Dockerfile's test stage stacks on TEST_BASE: this build's agent by
default, or a published image. It adds only the test tools (uv export
of the dev group into the image's environment), never reinstalling the
agent, so the suite tests the image as published.
- scripts/e2e.sh --image passes NISSE_E2E_MODEL into the container.
- The plan marks M1.5 done: v0.1.0's published image passed the whole e2e
suite (9 of 9, 11m37s), with the release's own tests over it.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
- Every PR: lint, unit tests across the runner's cores (pytest-xdist), 16
property walks from a seed that moves with the run number, integration
tests, and the end-to-end smoke tests (the slice, a conversation filed
as work, a question; `smoke` marker).
- A PR labelled e2e:full, a push to main, or a manual run: all 48 walks and
the whole end-to-end suite. Adding e2e:full to a PR runs it; any other
label runs nothing.
- A published release runs the full set before its image is built.
- The walk's seeds come from NISSE_WALK_SEEDS / NISSE_WALK_FIRST (48 from 0
by default).
Measured here: the unit tier on 2 workers took 3m49s against 6-8 minutes
serially.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
- Dockerfile: built with uv into /opt/nisse (installed, not editable),
run on python:3.12-slim with openssh-client as user nisse; stateless,
with state in a /var/lib/nisse-dev volume, config mounted at
/etc/nisse-dev/nisse-dev.yaml, secrets as environment variables, keys
mounted read-only. 202 MB. Its `test` stage adds the tests and their
tools over the installed package (UV_NO_SYNC, so `uv run` never swaps
in the source tree); the default target is the agent.
- scripts/e2e.sh --image REF runs the suite inside an image's test stage,
on its own sshd.
- CI builds both images on every PR (nothing pushed); release.yml pushes
gitea.dongar.net/dongar/nisse-dev:<tag> and :latest on a v* tag, with
the REGISTRY_TOKEN secret.
- scripts/bringup.sh builds the code under test into an image and runs it
in the systemd unit (root, $E2E mounted at its own path); `plan` prints
what it would run; the build-VM set-aside reads the projects: config.
Checked: the image refuses to start without a config, runs as uid 10001;
one e2e test passed from the test stage (2m45s, $0).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
dongarandClaude Opus 5.5
bot-claudeandClaude Opus 5.5
dev-agent