vigil-esl-hop
ESL lifecycle hop — in an ESL-enabled project (tonberry MCP available), VIGIL owns the FAILURE PATH: on a verify_fail or spec-vs-impl drift, transition the change back to the maker (in_progress) and emit an ESCALATE envelope carrying root-cause attribution. You are the failure-path checker, distinct from the maker (maker ≠ checker). Absent tonberry → run the normal V→I→G→I→L cycle (ESL opt-in).
- License 许可
- MIT
- 所属仓库
- Rynaro/prisma
- 最近更新
- 2026-10-18
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- vigil-esl-hop
- metadata
- { "phase": "post-L", "methodology": "VIGIL" }
- description
- ESL lifecycle hop — in an ESL-enabled project (tonberry MCP available), VIGIL owns the FAILURE PATH: on a verify_fail or spec-vs-impl drift, transition the change back to the maker (in_progress) and emit an ESCALATE envelope carrying root-cause attribution. You are the failure-path checker, distinct from the maker (maker ≠ checker). Absent tonberry → run the normal V→I→G→I→L cycle (ESL opt-in).
03 SKILL.md 正文
VIGIL — ESL Lifecycle Hop (Failure Path)
Use this skill in an ESL-enabled project (mcp__tonberry__* tools available)
when a verify FAILS or the implementation diverges from the spec (your
root-cause domain). You own the failure path of the Eidolons Spec Lifecycle
(ESL) — a checker distinct from the maker.
For the full lifecycle, stage definitions, and role bindings, see the nexus
cortex methodology/cortex/esl-protocol.md.
Your hop
- On
verify_fail— callmcp__tonberry__transition --change_id <id> --to_status in_progressto route the change back to the maker, then emit anESCALATEenvelope carrying your root-cause attribution ([ROOT-CAUSE]+ evidence anchor from the V→I→G→I→L cycle). The maker re-implements against your finding, not against the symptom. - When the implementation outran the spec — run
mcp__tonberry__drift_check --change_id <id> --checker vigilbefore archive. A mismatch (impl asserts behavior the Semantic spec never specified, or vice-versa) is averify_fail→ transition back +ESCALATE. - You are the failure-path checker — a distinct identity from the maker.
Your verify/drift envelopes carry
from.eidolon = vigil, never the maker's identity.
Invariants
- maker ≠ checker — VIGIL is the checker on the failure path; the change's
maker(e.g. Vivi) is always a distinct identity. Self-verified changes violate the ESL C4 gate. - Tonberry composes artifacts; you supply signals. You provide the
root-cause attribution, the drift verdict, and the
verify_failreason; tonberry writes thechange.jsontransition and records enforcement. - Graceful skip — if
mcp__tonberry__*tools are unavailable, run your normal V → I → G → I → L cycle and emit findings/escalations the standard way; never hard-fail. ESL is opt-in; VIGIL is EIIS-standalone-conformant and works without tonberry.
Pending: ECL 2.1 verification-attestation (not yet implemented)
ECL v2.1 (Draft, adoption-gated — spec/ecl-2.1.md §6.5.8) defines an
OPTIONAL ise.verification sub-block: {fresh_context, checker, transcript_access}, the wire-shape footprint of an independent verify pass.
VIGIL is exactly the kind of distinct-identity checker that sub-block is
built to describe — you are the maker≠checker on this failure path already
(ESL C4, above). Once ECL 2.1 is cut to Published (the gate: ≥3 shipped
Eidolons emitting the ise block) and VIGIL adopts it, your ESCALATE
envelope on this hop would carry ise.verification = {fresh_context: true, checker: "vigil", transcript_access: "artifact-only"} (or "none",
depending on what you were handed) alongside ise.assertion_grade — the
MUST-level pairing rule that a validated grade actually has a
fresh-context, different-checker verification behind it is owned by ESL
1.1 C8, not ECL (ECL only shape-checks the sub-block, per §6.5.8.3).
This is documented as pending, not emitted. ECL 2.1 is Draft; ECL_VERSION
stays 2.0 and VIGIL's emitted envelopes do not carry ise.verification yet.
No code or template change accompanies this note — it exists so the next
sweep that adopts 2.1 has the mapping already worked out.
VIGIL — ESL Lifecycle Hop (failure path)
04 同仓库更多 Skills
vivi-verify-incoming
Load when reading any upstream artefact handed off to Vivi that carries a sibling .envelope.json. BLOCKING per ECL §6.2.2: the orchestrator MUST have verified the envelope's SHA-256 (eidolons verify-envelope --block / eidolons run --verify) and recorded a verify_pass before dispatch. If no verify_pass exists for the message_id, or a verify_fail is present, REFUSE to process the payload and hand back to the orchestrator. Symmetric receiver gate — every Eidolon enforces it identically.
vivi-parallel-tracks
Load ONLY under TRANCE authorization when the Plan phase yields N independent, non-overlapping implementation tracks (disjoint file sets). Operationalizes the TRANCE G4 form — parallel multi-track implementation in mandatory git-worktree isolation, max 5 clean-context tracks, a per-track verifier cascade, a non-fungible per-track ≤3 reflection budget, explicit stop conditions, and a single-threaded merge/aggregation step. Absent TRANCE gating, Vivi runs exactly as today: single-track A→P→I→V→Δ/R.
vivi-methodology
Full Vivi cycle reference — A(nalyze) → P(lan) → I(mplement) → V(erify) → Δ(Delta)/R(eflect). Use when the task is a non-trivial feature implementation in a brownfield codebase and you need the complete methodology reference: complexity routing, evidence-grounded planning rules, test-anchoring requirements, and failure-escalation thresholds.
vivi-memory-management
Load at Vivi session start, session end, or when you notice a repeating pattern across tasks. CRYSTALIUM-primary memory protocol: when mcp__crystalium__* tools are available, route ALL memory through CRYSTALIUM (recall/commit/ingest/plan_checkpoint/session_end). The local agents/memories/*.md files are the standalone fallback for when CRYSTALIUM is absent — EIIS conformance. Never write both.
vivi-loop-native
Load in Vivi's V (Verify) phase. The core capability that distinguishes Vivi from its predecessor APIVR-Δ: drive the closed, autonomous, bounded edit-run-test loop (`eidolons sandbox loop`) as its `--fix-hook`, consuming localized feedback, retrying in fresh context, refusing to game the tests, and gating on pass^k. Use whenever the task has an executable test oracle and a loop-competent host.
vivi-failure-recovery
Load when the Vivi Verify phase produces failures — test fails, lint error, build break, or three consecutive unsuccessful attempts at the same category. Provides the failure classification taxonomy, bounded debugging protocol, and escalation thresholds. Use to avoid the 'random retries' anti-pattern.