ops(honcho): codify 3.0.9 host-ops runtime #777

Merged
pdurlej merged 1 commit from codex/honcho-309-host-ops-codify into main 2026-06-10 22:28:16 +02:00
Collaborator

Canary status: missing - fire canary 3+3 manually before merge

Canary Context Pack

Product story

Honcho is in active recovery/iteration and the repo must not describe a different runtime than production. After PR #776 landed 3.0.6 metadata, production was advanced again by a host-ops hotfix to home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z. This PR makes repo state catch up to the live, healthy host-ops state so future agents do not treat the current production state as unexplained drift.

What changed

  • Hardcoded Honcho API and deriver compose image refs to the live 3.0.9-gpt55-canonical-20260610T200313Z local image tag.
  • Hardcoded the Honcho sanitizer mount to /opt/pdurlej-platform/current/scripts/honcho/honcho_log_sanitize_exec.py, matching the active host-ops compose hotfix.
  • Updated honcho-api and honcho-deriver module metadata to the live image tag.
  • Added source/artifact metadata for honcho-deriver, matching its shared Honcho image.
  • Updated the module-catalog source-of-truth matrix for both Honcho services.
  • Added a metadata-only runtime evidence report.

Why it changed

Docker labels show the active production containers are currently composed from /opt/pdurlej-platform/runtime/host-ops/compose/..., not the release-root compose path used in the previous reconcile. The active host-ops compose file already hardcodes the 3.0.9 image and sanitizer path. Repo state was behind live state.

Files touched

  • compose/apps/compose.yaml
  • modules/honcho-api/module.yaml
  • modules/honcho-deriver/module.yaml
  • contexts/module-catalog/source-of-truth-matrix.md
  • state/reports/honcho-309-host-ops-reconcile-2026-06-10.md

Relevant context

  • PR #776 / #731 reconciled repo/runtime to 3.0.6, but a later host-ops hotfix moved production to 3.0.9.
  • Active runtime labels point to /opt/pdurlej-platform/runtime/host-ops/compose/apps/compose.yaml.
  • The host-ops pre-hotfix backup was /opt/pdurlej-platform/runtime/host-ops/compose/apps/compose.yaml.pre-honcho-gpt55-20260610T200313Z.

Runtime evidence

  • home-platform-honcho-api-1: running, healthy, home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z.
  • home-platform-honcho-deriver-1: running, healthy, home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z.
  • Both production containers and the local image tag resolve to image ID sha256:990ef700d2168ba38a2a30030118c80dc4e19e653d16d26a6394351c193f9338.
  • https://honcho.pdurlej.com/openapi.json returned HTTP 200.
  • Sanitized 15-minute warning/error scan showed no traceback, timeout, 429, 500, or 503 signal beyond normal uvicorn.error logger startup lines.
  • No raw Honcho messages, prompts, DB rows, payloads, auth headers, or secret values were inspected or recorded.

Known constraints

/opt/pdurlej-platform/runtime/host-ops/env/stack.env still has stale non-secret metadata (PROJECT_ROOT=/opt/vps-home-platform-infra, HONCHO_IMAGE=home-platform-honcho:3.0.3, HONCHO_IMAGE_REF=v3.0.3). The active compose file no longer uses those values for Honcho image or sanitizer mount. Cleaning those env keys is a separate runtime config mutation and is intentionally not done in this PR.

Explicit out-of-scope

  • Runtime mutation, container recreate, deploy, or merge/automerge.
  • Editing /opt/pdurlej-platform/runtime/host-ops/env/stack.env.
  • Database migrations, queue inspection, message/content inspection, secret changes, public exposure changes.
  • Deciding the long-term Honcho model/provider architecture.

Requested decision

Approve that repo compose/module metadata should match the live 3.0.9 host-ops hotfix state, with stale host-ops env cleanup tracked as a follow-up rather than bundled into this PR.

Merge blockers

  • Evidence that live 3.0.9 should be rolled back instead of codified.
  • Any secret or private Honcho payload content in the report.
  • Evidence that hardcoding the sanitizer path breaks the intended RS2000 host-ops contract.

Spec sources read

  • compose/apps/compose.yaml: canonical repo Honcho service definitions.
  • modules/honcho-api/module.yaml: current source/artifact/runtime metadata.
  • modules/honcho-deriver/module.yaml: current runtime metadata and missing artifact metadata.
  • contexts/module-catalog/source-of-truth-matrix.md: catalog source-of-truth summary.
  • state/reports/honcho-306-runtime-reconcile-2026-06-10.md: previous reconcile evidence and follow-up.
  • runbooks/honcho-ollama-gemma-switch.md: privacy/evidence rules for Honcho logs and metadata.

Verification

  • UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane platformctl validate modules/honcho-api/module.yaml --json
  • UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane platformctl validate modules/honcho-deriver/module.yaml --json
  • UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane python -c "import yaml, pathlib; yaml.safe_load(pathlib.Path('compose/apps/compose.yaml').read_text()); print('compose_yaml_parse=ok')"
  • git diff --cached --check
  • Ollama GLM advisory: no blocker; one risk noted around hardcoding future update friction, recorded as host-ops env cleanup follow-up.

Refs #731

Canary status: missing - fire canary 3+3 manually before merge ## Canary Context Pack ### Product story Honcho is in active recovery/iteration and the repo must not describe a different runtime than production. After PR #776 landed 3.0.6 metadata, production was advanced again by a host-ops hotfix to `home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z`. This PR makes repo state catch up to the live, healthy host-ops state so future agents do not treat the current production state as unexplained drift. ### What changed - Hardcoded Honcho API and deriver compose image refs to the live `3.0.9-gpt55-canonical-20260610T200313Z` local image tag. - Hardcoded the Honcho sanitizer mount to `/opt/pdurlej-platform/current/scripts/honcho/honcho_log_sanitize_exec.py`, matching the active host-ops compose hotfix. - Updated `honcho-api` and `honcho-deriver` module metadata to the live image tag. - Added source/artifact metadata for `honcho-deriver`, matching its shared Honcho image. - Updated the module-catalog source-of-truth matrix for both Honcho services. - Added a metadata-only runtime evidence report. ### Why it changed Docker labels show the active production containers are currently composed from `/opt/pdurlej-platform/runtime/host-ops/compose/...`, not the release-root compose path used in the previous reconcile. The active host-ops compose file already hardcodes the 3.0.9 image and sanitizer path. Repo state was behind live state. ### Files touched - `compose/apps/compose.yaml` - `modules/honcho-api/module.yaml` - `modules/honcho-deriver/module.yaml` - `contexts/module-catalog/source-of-truth-matrix.md` - `state/reports/honcho-309-host-ops-reconcile-2026-06-10.md` ### Relevant context - PR #776 / #731 reconciled repo/runtime to 3.0.6, but a later host-ops hotfix moved production to 3.0.9. - Active runtime labels point to `/opt/pdurlej-platform/runtime/host-ops/compose/apps/compose.yaml`. - The host-ops pre-hotfix backup was `/opt/pdurlej-platform/runtime/host-ops/compose/apps/compose.yaml.pre-honcho-gpt55-20260610T200313Z`. ### Runtime evidence - `home-platform-honcho-api-1`: running, healthy, `home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z`. - `home-platform-honcho-deriver-1`: running, healthy, `home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z`. - Both production containers and the local image tag resolve to image ID `sha256:990ef700d2168ba38a2a30030118c80dc4e19e653d16d26a6394351c193f9338`. - `https://honcho.pdurlej.com/openapi.json` returned HTTP 200. - Sanitized 15-minute warning/error scan showed no traceback, timeout, 429, 500, or 503 signal beyond normal `uvicorn.error` logger startup lines. - No raw Honcho messages, prompts, DB rows, payloads, auth headers, or secret values were inspected or recorded. ### Known constraints `/opt/pdurlej-platform/runtime/host-ops/env/stack.env` still has stale non-secret metadata (`PROJECT_ROOT=/opt/vps-home-platform-infra`, `HONCHO_IMAGE=home-platform-honcho:3.0.3`, `HONCHO_IMAGE_REF=v3.0.3`). The active compose file no longer uses those values for Honcho image or sanitizer mount. Cleaning those env keys is a separate runtime config mutation and is intentionally not done in this PR. ### Explicit out-of-scope - Runtime mutation, container recreate, deploy, or merge/automerge. - Editing `/opt/pdurlej-platform/runtime/host-ops/env/stack.env`. - Database migrations, queue inspection, message/content inspection, secret changes, public exposure changes. - Deciding the long-term Honcho model/provider architecture. ### Requested decision Approve that repo compose/module metadata should match the live 3.0.9 host-ops hotfix state, with stale host-ops env cleanup tracked as a follow-up rather than bundled into this PR. ### Merge blockers - Evidence that live 3.0.9 should be rolled back instead of codified. - Any secret or private Honcho payload content in the report. - Evidence that hardcoding the sanitizer path breaks the intended RS2000 host-ops contract. ## Spec sources read - `compose/apps/compose.yaml`: canonical repo Honcho service definitions. - `modules/honcho-api/module.yaml`: current source/artifact/runtime metadata. - `modules/honcho-deriver/module.yaml`: current runtime metadata and missing artifact metadata. - `contexts/module-catalog/source-of-truth-matrix.md`: catalog source-of-truth summary. - `state/reports/honcho-306-runtime-reconcile-2026-06-10.md`: previous reconcile evidence and follow-up. - `runbooks/honcho-ollama-gemma-switch.md`: privacy/evidence rules for Honcho logs and metadata. ## Verification - `UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane platformctl validate modules/honcho-api/module.yaml --json` - `UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane platformctl validate modules/honcho-deriver/module.yaml --json` - `UV_CACHE_DIR=/private/tmp/uv-cache uv run --project control-plane python -c "import yaml, pathlib; yaml.safe_load(pathlib.Path('compose/apps/compose.yaml').read_text()); print('compose_yaml_parse=ok')"` - `git diff --cached --check` - Ollama GLM advisory: no blocker; one risk noted around hardcoding future update friction, recorded as host-ops env cleanup follow-up. Refs #731
ops(honcho): codify 3.0.9 host-ops runtime
All checks were successful
canary-required / collect-diff (pull_request) Successful in 4s
platformctl plan / auto-apply scope (pull_request) Successful in 20s
base-is-main / guard (pull_request) Successful in 1s
patchwarden-pr-sanity / collect-diff (pull_request) Successful in 4s
canary-required / canary (pull_request) Has been skipped
patchwarden-client-dry-run / collect-diff (pull_request) Successful in 4s
patchwarden-client-dry-run / dry-run (pull_request) Successful in 18s
patchwarden-pr-sanity / sanity (pull_request) Successful in 3m35s
64c39145a9
pdurlej deleted branch codex/honcho-309-host-ops-codify 2026-06-10 22:28:16 +02:00
First-time contributor

Patchwarden PR sanity

  • Status: advisory_findings
  • PR: 777
  • Commit: 64c39145a90764017f6b5ff21ecec34279ac3c2e
  • Security-sensitive label: missing
  • Authority: advisory model review plus deterministic blockers only
  • 3+3 canary: still alive; this does not replace it

Deterministic findings

No deterministic findings.

Model reviewers

global-glm / glm-5.1:cloud

  • Status: ok
  • Verdict: OK
  • Findings: none

global-deepseek / deepseek-v4-pro:cloud

  • Status: ok
  • Verdict: OK
  • Findings: none

redteam / kimi-k2.6:cloud

  • Status: ok

  • Verdict: NOT_OK

  • blocker Merge proposed with missing canary verification

    • Evidence: PR description header explicitly states 'Canary status: missing - fire canary 3+3 manually before merge', indicating the required pre-merge validation protocol has not been executed.
    • Next: Execute the 3+3 canary protocol, validate health/metrics, and update the PR header to reflect passing canary status before merging.
  • high Host-side sanitizer script bind-mounted into containers

    • Evidence: PR description admits 'Hardcoded the Honcho sanitizer mount to /opt/pdurlej-platform/current/scripts/honcho/honcho_log_sanitize_exec.py' in compose/apps/compose.yaml. Mounting a mutable host executable into containers creates a code-injecti
    • Next: Bake the sanitizer script into the container image at build time instead of bind-mounting from the host. If a host mount is mandatory, mount it read-only with an immutable base path and add an integrity checksum verified at container start.
  • high Compose hardcodes mutable local image tag instead of digest

    • Evidence: PR description states compose image refs are set to tag 'home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z' while separately noting the resolved digest is sha256:990ef700d2168ba38a2a30030118c80dc4e19e653d16d26a6394351c193f9338. Lo
    • Next: Pin the image reference by digest in compose/apps/compose.yaml (e.g., image@sha256:990ef700d216...) to guarantee immutable supply-chain integrity.

Policy notes

  • GLM 5.1 + DeepSeek V4 Pro are the operator-required model mix for this bot.
  • Optional red-team model is enabled only when PLATFORMCTL_PR_SANITY_REDTEAM_MODEL is configured.
  • Auto-merge is not enabled here.
<!-- patchwarden-pr-sanity:pdurlej/platform:PR-777 --> # Patchwarden PR sanity - Status: `advisory_findings` - PR: `777` - Commit: `64c39145a90764017f6b5ff21ecec34279ac3c2e` - Security-sensitive label: `missing` - Authority: advisory model review plus deterministic blockers only - 3+3 canary: still alive; this does not replace it ## Deterministic findings No deterministic findings. ## Model reviewers ### `global-glm` / `glm-5.1:cloud` - Status: `ok` - Verdict: `OK` - Findings: none ### `global-deepseek` / `deepseek-v4-pro:cloud` - Status: `ok` - Verdict: `OK` - Findings: none ### `redteam` / `kimi-k2.6:cloud` - Status: `ok` - Verdict: `NOT_OK` - **`blocker`** Merge proposed with missing canary verification - Evidence: `PR description header explicitly states 'Canary status: missing - fire canary 3+3 manually before merge', indicating the required pre-merge validation protocol has not been executed.` - Next: Execute the 3+3 canary protocol, validate health/metrics, and update the PR header to reflect passing canary status before merging. - **`high`** Host-side sanitizer script bind-mounted into containers - Evidence: `PR description admits 'Hardcoded the Honcho sanitizer mount to /opt/pdurlej-platform/current/scripts/honcho/honcho_log_sanitize_exec.py' in compose/apps/compose.yaml. Mounting a mutable host executable into containers creates a code-injecti` - Next: Bake the sanitizer script into the container image at build time instead of bind-mounting from the host. If a host mount is mandatory, mount it read-only with an immutable base path and add an integrity checksum verified at container start. - **`high`** Compose hardcodes mutable local image tag instead of digest - Evidence: `PR description states compose image refs are set to tag 'home-platform-honcho:3.0.9-gpt55-canonical-20260610T200313Z' while separately noting the resolved digest is sha256:990ef700d2168ba38a2a30030118c80dc4e19e653d16d26a6394351c193f9338. Lo` - Next: Pin the image reference by digest in compose/apps/compose.yaml (e.g., image@sha256:990ef700d216...) to guarantee immutable supply-chain integrity. ## Policy notes - GLM 5.1 + DeepSeek V4 Pro are the operator-required model mix for this bot. - Optional red-team model is enabled only when `PLATFORMCTL_PR_SANITY_REDTEAM_MODEL` is configured. - Auto-merge is not enabled here.
Sign in to join this conversation.
No reviewers
No labels
W6d-automerge-calibration
agent/claude-code
agent/codex
agent/hermes
agent/iskra
agent/ollama
agent/patchwarden
automerge-candidate
class/security-sensitive
cutover-gate
dependency/blocked
dependency/blocks-others
dependency/cross-repo
dependency/needs-confirmation
domain:agents
domain:ci
domain:docs
domain:forgejo
domain:infra
domain:memory
domain:runtime
domain:signal
domain:ux
flow/architecture
flow/blocked
flow/deployed
flow/done
flow/implementation
flow/intake
flow/maintained
flow/observed
flow/ready
flow/refining
flow/retired
flow/review
iterating
judge/codex-candidate
judge/hermes-candidate
judge/low-confidence
judge/needs-refinement
judge/operator-needed
judge/p0
judge/p1
judge/p2
judge/p3
judge/park
judge/patchwarden-candidate
judge/stale-priority
kind/adr
kind/bug
kind/chore
kind/feature
kind/infra
kind/ops
kind/refactor
kind/research
large-impact
merge/auto
merge/manual
merge/manual-dependency-conflict
merge/manual-failing-tests
merge/manual-merge-conflict
merge/manual-missing-review
merge/manual-operator-preference
merge/manual-red-zone
merge/manual-security-sensitive
merge/manual-unclear-scope
merge/manual-unknown
meta
mode:operator-only
mode:patchwarden-iskra-approved
mode:safe-auto
needs-operator-decision
needs-triage
not-ready
observed/erroring
observed/needs-followup
observed/pending
observed/retire-candidate
observed/unused
observed/used
operator-emotional
owner-attention
phase/02
phase/03
priority:p0
priority:p1
priority:p2
priority:p3
proposed
ready-for-agent
ready-for-operator
recovery
review:claude-reviewed
review:codex-reviewed
review:dziadek-reviewed
review:needs-human
risk/exposure
risk/process
risk/product
risk/runtime
safety:external-write
safety:no-prod-mutation
safety:prod-impact
safety:secret-touch
size/large
size/medium
size/small
size/tiny
size/unknown
source/adr
source/agent-generated
source/manual
source/operator-chat
source/voice-note
status:blocked
status:codex-ready
status:merged:pending-evidence
status:needs-evidence
status:operator-needed
status:parked
tier/full
tier/lite
tier/stacked
tier:0-platform-substrate
tier:1-iskra-value-layer
tier:2-tools-products-modules
type:bug
type:chore
type:docs
type:feat
type:policy
type:research
No milestone
No project
No assignees
2 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
pdurlej/platform!777
No description provided.