Converge native Aguard on operator hosts #635
Labels
No labels
burndown-2026-06
autonomy
async-consult
autonomy
epic
autonomy
headless
autonomy
live-collab
coherence-core
priority
P0
priority
P1
priority
P2
priority
P3
priority
P4
qa-fixture
role/ai
role/creator
role/design
role/director
role/engineer
role/exec
role/human
role/ops
role/qa
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
coilyco-flight-deck/infrastructure#635
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Problem
The AOS operator CLI cutover makes
aguardthe native operator surface as well as the container surface. Full dev-base already installs Aguard, but long-lived operator hosts need it through normal fleet convergence.AOS authoring and portable packaging are tracked in agentic-os#728. The Ward-side removal of generated operator leaves is tracked in ward#1548.
Ownership
Infrastructure owns persistent host rollout. Do not put fleet mutation in Ward container assets, AOS package post-install hooks, or ad hoc shell startup code.
Change
aosandaguardcommands.ward ops, ward-kdl, or a source checkout for AOS-owned operator leaves once the dependencies land.Acceptance
aguardonPATH.aos --version,aguard --version, andaguard --helpsucceed from a clean login environment.Dependency
Implementation may prepare the convergence shape now, but rollout cannot complete until agentic-os#728 publishes the native Aguard artifact.
WARD-WORKFLOW: #636
details
review: skipped. The in-container review gate was intentionally skipped because the temporary Ward default is pending brokered QA.
workflow: pull-request; review summary: skipped pending brokered QA
Retrospective: The staged dependency boundary felt clear. The only surprise was the documentation-size and YAML-comment policy, both resolved before commit.
Confidence: high for the prepared Ansible shape. Follow-up: enable the declared paired release after agentic-os#728 publishes it.
WARD-WORKFLOW: reservation-released
release details
Run finished with
WARD-WORKFLOW: https://forgejo.coilysiren.me/coilyco-flight-deck/infrastructure/pulls/636.ward container reapreleased containerengineer-codex-infrastructure-635(--harness codex): the terminal outcome supersedes the reservation, so a later redispatch no longer needs--override-reservation.Outcome summary: #636
— Codex, via
ward agentIncident impact observed
During the recovery recorded in infrastructure#642, Scoop upgraded Ward from 0.800.0 to 0.805.0 while the operator was working. The new Ward binary no longer exposed the old
opssurface, and nativeaguardwas not yet available on PATH.The operator completed the guarded Forgejo credential lookup with the still-installed Ward 0.800.0 binary. No secret was printed or written to disk.
This was a live incident-response availability gap, not only stale documentation. The acceptance criteria here should include an upgrade ordering guarantee: native Aguard must be discoverable from a clean login environment before fleet convergence removes the last working operator surface.