Compare commits

...
13 Commits
Author SHA1 Message Date
dev-agent d407db72c2 Merge pull request 'Revise the deployment model: an agent host and tool VMs' (#90) from deployment-model into main
CI / checks (push) Successful in 1m24s
2026-10-02 19:36:33 +00:00
dongarandClaude Opus 5.5 2187c24d0d Address review comments on the architecture
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m22s
Name the action classes harmless, contained and dangerous instead of
numbering them; say what escalation doesn't depend on; say when a work
item's Matrix thread starts and how threads route commands; add open
questions for the state store's engine, ops issuing credentials, and
releases.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 19:33:17 +00:00
dongarandClaude Opus 5.5 8c38444b88 Close five gaps the loop model found
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m18s
- abandon closes the item's open PRs
- a PR the operator closed is an answer, never reopened by a retry
- a review fix for a PR that merged meanwhile goes out as a new PR
- redoing a milestone after a closed PR gets a new branch and PR
- a record issue carries its work item's ID from creation

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 19:24:54 +00:00
dongarandClaude Opus 5.5 5839091d65 Define projects as configuration entries (D8); note the agent host's recovery
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m19s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:53:45 +00:00
dongarandClaude Opus 5.5 4ca8e1f550 Drop Vikunja: Gitea issues are the one tracker for every agent
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m17s
Ops work lives in one infrastructure repository for both sites, with
site labels; intake, labels and dependencies work as for nisse-dev, and
agent-to-agent requests are cross-repo issues.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:52:18 +00:00
dongarandClaude Opus 5.5 6e4dca040c Condense the architecture; move mechanism detail to the Development Plan
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m18s
Each rule is stated once in its home section and referred to by ID
elsewhere. No decision added or removed; implementation detail moved to
Development-Plan.md section 6, keyed by architecture section. P8 folded
into P1, P5 and X1; the Appendix about the old spec dropped. Section
IDs are unchanged.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:43:16 +00:00
dongarandClaude Opus 5.5 d77c142488 Accept a plain-language go-ahead, announced and revision-bound
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m16s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:35:00 +00:00
dongarandClaude Opus 5.5 349ec794f2 Fix inconsistencies and gaps found re-reading the architecture
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m27s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:30:32 +00:00
dongarandClaude Opus 5.5 acf7ce191c Let the agent file issues from conversation (later addition)
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m17s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:26:58 +00:00
dongarandClaude Opus 5.5 cd2ff6d769 Write the agreed dev workflow into the architecture
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m19s
Issues assigned to dev-agent are the intake; the agent restates its
understanding and either proceeds (bug label) or waits for a go-ahead.
Labels steer and are never required. Build plans and ARCHITECTURE.md
are working documents reviewed in PRs, not gates; scope is the issue
and its conversation. Work items can be blocked on Gitea dependencies,
each gets a Matrix thread, milestones are started and finished by
control tools, and the next milestone waits for the previous merge.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 18:26:02 +00:00
dongarandClaude Opus 5.5 3e7edce6dc Make the vault's git history optional
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m20s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 17:37:45 +00:00
dongarandClaude Opus 5.5 aa97309a0c Replace Outline with an Obsidian-format knowledge vault on the agent host
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m21s
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 17:30:54 +00:00
dongarandClaude Opus 5.5 230b927ebe Revise the deployment model: an agent host and tool VMs
CI / checks (push) Skipped
CI / checks (pull_request) Successful in 1m20s
Agents run as containers on one stable agent host that keeps their
state and secrets and never runs a command a model chose. Model-chosen
commands run on disposable tool VMs over SSH: a build VM per project
for nisse-dev, its own tool VM for nisse-ops. Secrets are a set per
agent, handed out per operation by the harness and redacted from tool
output. MCP servers stay on the agent host as the one exception.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE
2026-10-02 16:35:55 +00:00
2 changed files with 773 additions and 792 deletions

No files matched your search

+719 -792
View File
File diff suppressed because it is too large. Load diff
+54
View File
@@ -309,3 +309,57 @@ All 19 open issues, as of 2026-09-30:
| #79 ending projects and milestones | **Superseded** | M2.5 | | #79 ending projects and milestones | **Superseded** | M2.5 |
| #80 build artifacts in PRs | Milestone | M1.4 | | #80 build artifacts in PRs | Milestone | M1.4 |
| #81 no-change milestones | **Superseded** | M2.5 | | #81 no-change milestones | **Superseded** | M2.5 |
## 6. Implementation notes
Mechanism detail moved out of `docs/Architecture.md` in PR #90, which keeps only
decisions and their reasons. Keyed by the architecture section each note
implements; a milestone touching that section uses the note, and changes it here
if the implementation differs.
- **L1 switching:** the grace period before switching away from an item waiting
on the operator defaults to 10 minutes. When the driver chooses the next ready
item and the choice fails or is invalid, fall back to the oldest.
- **L3 revision stamping:** in Matrix, a reply is stamped with the revision whose
message had been sent when it arrived; on a Gitea issue, with the latest
proposal comment older than the reply.
- **L5 transport:** the `Transport` interface is send, ask and receive. Intake
backoff is exponential, capped at a few minutes.
- **L6 fitting order:** first shorten old tool results to their first and last
lines, then drop the oldest events with a note saying so; never the system
prompt, the current plan or the most recent turns. View labels:
`[orchestrator]`, `[operator, Matrix]`, `[operator, comment on PR #N]`,
`[helper report]`, `[executor]` narration.
- **L7 rendering:** content is stored as bytes or local paths and uploaded or
inlined at render time, never as provider handles. Token budgeting and
prompt-cache breakpoints are render-time concerns, never stored. Compaction
summaries are first-class events so a model change never inherits opaquely
compacted history. A reasoning "session" is a delegation, including one resumed
after a crash.
- **L8 state store:** SQLite in WAL mode, each change committed and fsynced before
the loop continues. A delegation's cut is clean when its last event is a
completed result, its branch is at the expected commit, its caps are unbreached
and external truth matches the log. "Anything else" offers continue, re-plan or
abandon.
- **L9 local models:** request `tool_choice` only of models whose server declares
it can enforce it; otherwise the loop enforces the contract (an explicit
`finish`, prose treated as a violation). Briefings use the capability profile,
never a model name or "weaker model".
- **L10 vault:** atomic writes are write-then-rename. Obsidian's settings folder
(and `.git`, if history is kept) are not synced. A failed write is retried, or
recorded and reconciled later.
- **L12:** secrets reach a command by environment or stdin. Headless agents get an
Authelia single-factor exception scoped by source address, group and target.
Logs are shipped to the host journal, not kept in containers.
- **D3 preflight:** a project's needs are read from its files (`Cargo.toml`,
`pyproject.toml`, `package.json`) or a declaration in the repository.
- **D4 publishing:** standard excludes are configured in each clone, and only what
the work produced is committed. A push hands the key over stdin to a
short-lived `ssh-agent` on the build VM, for that one push.
- **D7:** test services use ports 18000–18099, so opening access later is a
firewall change. The inference host is behind LiteLLM, so it is not a second
egress exception.
- **X4:** the agent host's state dataset is ZFS, snapshotted on a schedule and on
each approval. Tool VM templates use Terraform's `bpg/proxmox` provider with
cloud-init. Containers restart unless stopped.
- **X5:** the local inference host is llama-swap on an RTX 4090 machine.