Propose agent execution plane v1

This commit is contained in:
m
2026-08-22 11:34:26 +03:00
parent 339f6a6770
commit 82ee6e5ba0
3 changed files with 122 additions and 2 deletions
+2 -2
View File
@@ -1,5 +1,5 @@
{
"task_file": "dev/tasks/DEV-000103-epistemics-truth-v1-proposal.json",
"task_id": "DEV-000103",
"task_file": "dev/tasks/DEV-000104-agent-execution-plane-proposal.json",
"task_id": "DEV-000104",
"task_revision": 1
}
@@ -0,0 +1,23 @@
{
"task_id": "DEV-000104",
"task_revision": 1,
"architecture_revision": "ADR-0018",
"base_repo_revision": "339f6a6",
"risk_tier": "E",
"intent": "Prepare the proposed ADR-0020 Agent Execution Plane boundary for review without implementing dispatcher, worker, schema, callback or deployment changes.",
"requirement_ids": ["WP-300", "Durable dispatch lease heartbeat", "Independent verification and truthful UNKNOWN outcome"],
"allowed_change_scope": {"modules": ["work-execution-governance"], "paths": ["docs/proposals/ADR-0020-agent-execution-plane-v1.md", "dev/ACTIVE_TASK.json", "dev/tasks/DEV-000104-agent-execution-plane-proposal.json"]},
"forbidden_change_scope": {"paths": ["architecture/**", "contracts/**", "database/**", "migrations/**", "modules/**", "cmd/**", "runtime/**", "adapters/**", "deploy/**", ".gitea/**"]},
"allowed_architecture_delta": {"new_cross_module_dependencies": 0, "new_canonical_owners": 0, "new_privileges": 0, "new_third_party_dependencies": 0, "new_database_owner_roles": 0},
"contracts": {"consume": ["ADR-0017 General Agent Delegation and Codex Bridge V1", "Proposed ADR-0019 Epistemics Truth V1", "Core semantic model Work/Capability/Artifact/Continuity boundaries"], "may_change": ["Proposed ADR-0020 only; no accepted contract changes"]},
"semantic_owner": "work-execution-plane-proposal-only",
"canonical_state": [],
"state_transitions": ["dispatch -> leased -> running -> result -> independently verified", "lease expiry/callback uncertainty -> UNKNOWN or RECOVERING", "cancel request -> acknowledged cancellation or typed unknown"],
"side_effects_and_privileges": ["No runtime side effects, external mutation, connector access, credential access or authority expansion"],
"persistence_and_migrations": ["No schema, migration, queue, worker or lease implementation changes"],
"failure_retry_restart": ["Retry creates a new run and preserves history; late or duplicate callbacks cannot become authoritative"],
"assumptions": ["Work remains canonical owner of durable delegation execution", "Artifact remains canonical owner of attestation metadata", "Production deployment remains outside this proposal"],
"acceptance_scenarios": ["Worker identity and dispatch idempotency are explicit", "Lease fencing and heartbeat expiry are explicit", "Callback authentication/replay handling is explicit", "Cancellation/retry/UNKNOWN semantics preserve history", "Independent verification gates Work completion", "No implementation or registry delta is introduced"],
"required_tests": ["go test ./...", "go vet ./...", "go run ./cmd/devctl preflight -active dev/ACTIVE_TASK.json", "go run ./cmd/devctl check-diff -active dev/ACTIVE_TASK.json", "go run ./cmd/devctl check-architecture", "go run ./cmd/devctl/independent", "git diff --check", "proposal scope and required-section review"],
"terminal_gap_states": ["BLOCKED_BY_ARCHITECTURE_GAP", "BLOCKED_BY_SPEC_GAP", "BLOCKED_BY_CONFLICT"]
}
@@ -0,0 +1,97 @@
# Proposal ADR-0020 — Agent Execution Plane V1
Status: Proposed
Task: DEV-000104
## Problem and evidence
ADR-0017 defines durable delegation semantics, but the audit found no
automatic dispatcher/worker plane, lease/heartbeat lifecycle, callback
authentication or independent verification boundary. A worker must be able
to stop, restart and resume without turning a missing callback into success or
silently duplicating external work.
## Decision proposed
1. Keep `work` as the canonical owner of `Delegation`, `AgentRun`,
`AgentCheckpoint`, `AgentResult` and verification gating. `capability`
owns executor-neutral agent implementations; adapters own transport and
process details.
2. Every dispatch has a stable `dispatch_id`, delegation id, run id,
implementation id, exact source/worktree identity, bounded instruction
artifact reference, requested capability set, authority/credential refs,
deadline and idempotency key. The derived instruction must preserve
OriginalHumanIntent, inferred goal, added context and verification plan as
separate provenance fields.
3. A worker must acquire a durable lease before execution. Lease ownership is
a fencing token, not an in-memory mutex. Heartbeats extend the lease only
for the same run and fencing token; an expired lease makes late callbacks
non-authoritative.
4. Dispatch and callback endpoints authenticate the worker identity and bind
every message to delegation, run, dispatch, implementation and fencing
token. Forged, replayed, out-of-order or duplicate callbacks are rejected
or recorded as non-authoritative evidence. Idempotent acknowledgement is
allowed only for the exact previously accepted event.
5. Checkpoints and artifacts are content-addressed metadata owned by
`artifact`, with typed refs from Work. Artifact attestation records source
revision/worktree, producer, digest, test command/result and capture time;
an un-attested path or log is not completion evidence.
6. Cancellation is a durable Work decision. The dispatcher sends cancellation
to the leased worker, records acknowledgement or timeout, and never erases
history. A worker that cannot prove cancellation remains `UNKNOWN` or
`RECOVERING` until reconciled.
7. Retry is policy-driven and distinguishes retryable transport/worker loss,
execution-window exhaustion, human blocker, invalid result and unknown
external outcome. Retry creates a new run under the same delegation and
preserves prior checkpoints/results; it never reuses an idempotency key for
an unproven external mutation.
8. A run result cannot complete Work directly. Independent verification must
consume the declared artifacts and tests through a separate verifier
identity and produce a typed verification result. Verification failure,
missing evidence and verifier unavailability remain distinct states.
9. Secrets are resolved only through approved `CredentialRef` boundaries and
are not copied into instructions, callbacks, artifacts, logs or task
packets. Worker startup has no ambient credential inheritance.
10. The plane is durable and restartable: dispatch intent, lease, heartbeat,
cancellation, checkpoint, result, verification and reconciliation events
are persisted before their projections claim progress.
## Lifecycle proposal
Delegation remains the higher-level human/work intent. A run progresses through
`DISPATCHED`, `LEASED`, `RUNNING`, `CHECKPOINTED`, `SUCCEEDED`, `FAILED`,
`CANCEL_REQUESTED`, `CANCELLED`, `EXHAUSTED`, `WAITING_HUMAN`, `UNKNOWN` or
`RECOVERING`, subject to explicit transition guards. `SUCCEEDED` means only
that the worker returned an accepted result; Work completion additionally
requires independent verification. Lease expiry and missing heartbeat never
become success.
## Alternatives and rationale
- An in-memory queue or goroutine was rejected because it loses dispatch state
on restart and violates durable timer/lease requirements.
- A single generic `AgentManager` was rejected because it collapses Work,
Capability, Artifact and Continuity ownership.
- Trusting worker-reported success was rejected because the worker and
verifier must remain separate authorities.
- Reusing the same run on retry was rejected because it makes duplicate
external effects and late callbacks indistinguishable.
## Boundaries, consequences and non-goals
No new canonical owner, provider, deployment authority, database migration,
Codex-specific domain contract or automatic self-modification is proposed.
The design enables restartable external work and truthful failure handling,
but requires durable event ordering, fencing, artifact storage and verifier
capacity before WP-301 can claim production readiness.
## Migration, rollback and revisit triggers
Implementation must begin behind an opt-in dispatch path with a dry-run
verifier and deterministic fake worker. Existing delegation records remain
readable; unknown legacy RUNNING records are classified by WP-302 rather than
silently resumed. Rollback disables new dispatch and drains/recovers leases;
history and artifacts remain. Revisit if a worker needs authority beyond its
declared capability, if a second transport changes callback guarantees, or if
artifact attestation cannot be made independent of the worker.