Run any Skill in Manus with one click

vestige-trace-replay

Stars5

Forks0

UpdatedMay 8, 2026 at 19:41

Use this skill to re-run a stored trace against the current memory store and diff the results against the original — answering "would the agent get the same answer today?". Fire when the user says "re-run trace_X", "replay that search", "diff trace_X against current state", or "did the answers drift?". Read-only: writes a NEW query_events row tagged with replay_of; never mutates the original trace or any memory. provider_match=false flags embedding drift.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

conorluddy

conorluddy/Vestige

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Software Quality Assurance Analysts and TestersComputer and Mathematical Occupations·SOC 15-1253

SKILL.md

readonly

name

vestige-trace-replay

description

Replay a stored query trace

Re-run a stored query trace against the current store and current embedding provider. Produces an explicit diff — added/removed result IDs, score deltas — and flags whether the provider and corpus match the original. This is the "did memory drift?" command.

When to fire

Drift detection. After a major memory session (many records, approvals, or forgets) — confirm that prior recall results are stable.
Provider change. Switching embedding providers or models — replay key traces to quantify the impact.
Corpus growth audit. New memories were captured since a trace was written — replay shows what's been added to the result set.
Regression check. Before a PR that touches recall logic — replay a set of traces and surface score deltas.

Replay is read-only with respect to memories. It writes a new query_events row tagged caller=cli (or mcp) with replay_of: <original_trace_id> in params_json, making the replay chain inspectable. The original trace is never modified.

How to invoke

vestige trace replay <trace_id> [--json]

Accepts a trace_<ULID> — obtain one from vestige trace (list mode).
provider_match=false in the output signals that the current embedding provider or model differs from the one in the original trace. When this happens the mode may fall back to lexical and mode_fallback=true is set.

vestige trace replay trace_01JWXXXXXXXXXXXXXXXXXX           # text diff
vestige trace replay trace_01JWXXXXXXXXXXXXXXXXXX --json    # JSON diff

After invocation

{
  "trace_id": "trace_01JWXXXXXXXXXXXXXXXXXX",
  "original": {
    "result_ids": ["mem_01JWXXXXXXXXXXXXXXXXXX", "mem_01HVXXXXXXXXXXXXXXXXXX"],
    "scores":     [0.83, 0.61]
  },
  "current": {
    "result_ids": ["mem_01JWXXXXXXXXXXXXXXXXXX", "mem_01HVXXXXXXXXXXXXXXXXXX", "mem_01KAXXXXXXXXXXXXXXXXXX"],
    "scores":     [0.83, 0.61, 0.55]
  },
  "diff": {
    "added":         ["mem_01KAXXXXXXXXXXXXXXXXXX"],
    "removed":       [],
    "score_changes": []
  },
  "provider_match":  true,
  "mode_fallback":   false,
  "corpus_drift":    1,
  "replay_trace_id": "trace_01KBXXXXXXXXXXXXXXXXXX"
}

corpus_drift is the net change in result count. replay_trace_id is the ID of the new query_events row written for this replay run — inspect it with vestige-trace-show <replay_trace_id>.

More from this repository

same repository

vestige-scan-sessions

conorluddy/Vestige

Mine recent local coding-agent transcripts for memories worth keeping, using the `vestige_scan_sessions` MCP tool. Fire at session start (to catch up on what happened in prior sessions), when the user says "scan my sessions", "what did we work on recently?", "catch up from my transcripts", "ingest recent sessions", "mine my history", or when picking up a project after a pause. The tool hands you redacted, cursor-advanced turns from this project's Claude Code and Codex transcripts; you extract decisions/notes/preferences/questions inline and file each via `vestige_propose_candidate`. Opt-in — requires `[mcp] allow_scan_sessions = true` in `.vestige/config.toml`. Candidates land in the assimilation inbox for human review, never as durable memories directly.

2026-06-145

vestige-sources

conorluddy/Vestige

Use this skill when you need the raw source receipts for a memory or candidate — the typed evidence rows, not the narrative walk. Fire when the user says "what sources back mem_X?", "show me the source receipts", "list the sources for mem_X", or "filter sources by kind agent_session". Use `vestige-why` when you want the full provenance narrative including events; use this skill when you want the tabular source index only.

2026-05-085

vestige-trace-list

conorluddy/Vestige

Use this skill when you want to audit what queries the agent or CLI has run against project memory. Fire when the user says "what queries did the agent run?", "show recent trace events", "did anyone search for X recently?", or "list traces by caller mcp". Every vestige search, recall, expand, and context call writes a query_events row automatically — this skill surfaces them. Use vestige-trace-show to expand a single trace.

2026-05-085

vestige-trace-show

conorluddy/Vestige

Use this skill when you want to expand a specific trace to its full detail — the query text, mode requested vs resolved, provider, result IDs, and scores. Fire when the user says "expand trace_X", "show me what trace_X returned", or "what did that search find?". Use vestige-trace-list first to find the trace_id, then this skill to inspect one in full.

2026-05-085

vestige-why

conorluddy/Vestige

Use this skill when you want to know where a memory came from, need to explain its provenance, or must audit what was captured and when. Fire when the user says "where did this come from?", "explain that memory", "why do we have mem_X?", "show provenance for X", "what's the source of mem_X?", or "trace this memory back". Works for both memories and pre-approval candidates. Soft-deleted memories are included — provenance is always inspectable.

2026-05-085

vestige-review

conorluddy/Vestige

Review Vestige code against this repo's CODESTYLE.md, CLAUDE.md, and PRD — naming, progressive disclosure, error handling, agentic patterns, and the seven hard architectural rules (soft-delete only, project-scope boundary, immutable migrations, one-way crate deps, newtype IDs, MCP intent-not-mechanics, no daemon). Use when the user says "vestige-review", "review this", "review my branch", "check my code", "look over this PR/diff", "roast this", "is this ready to merge", or after finishing work in any Vestige crate. Prefer over the global code-review skill inside this repo — it layers Vestige-specific rules on top of the style guide and produces findings grouped by severity (blocker / major / minor / nit) with file:line citations and concrete fixes.

2026-05-015

name

vestige-trace-replay

description

Replay a stored query trace

When to fire

Drift detection. After a major memory session (many records, approvals, or forgets) — confirm that prior recall results are stable.
Provider change. Switching embedding providers or models — replay key traces to quantify the impact.
Corpus growth audit. New memories were captured since a trace was written — replay shows what's been added to the result set.
Regression check. Before a PR that touches recall logic — replay a set of traces and surface score deltas.

How to invoke

vestige trace replay <trace_id> [--json]

Accepts a trace_<ULID> — obtain one from vestige trace (list mode).
provider_match=false in the output signals that the current embedding provider or model differs from the one in the original trace. When this happens the mode may fall back to lexical and mode_fallback=true is set.

vestige trace replay trace_01JWXXXXXXXXXXXXXXXXXX           # text diff
vestige trace replay trace_01JWXXXXXXXXXXXXXXXXXX --json    # JSON diff

After invocation

{
  "trace_id": "trace_01JWXXXXXXXXXXXXXXXXXX",
  "original": {
    "result_ids": ["mem_01JWXXXXXXXXXXXXXXXXXX", "mem_01HVXXXXXXXXXXXXXXXXXX"],
    "scores":     [0.83, 0.61]
  },
  "current": {
    "result_ids": ["mem_01JWXXXXXXXXXXXXXXXXXX", "mem_01HVXXXXXXXXXXXXXXXXXX", "mem_01KAXXXXXXXXXXXXXXXXXX"],
    "scores":     [0.83, 0.61, 0.55]
  },
  "diff": {
    "added":         ["mem_01KAXXXXXXXXXXXXXXXXXX"],
    "removed":       [],
    "score_changes": []
  },
  "provider_match":  true,
  "mode_fallback":   false,
  "corpus_drift":    1,
  "replay_trace_id": "trace_01KBXXXXXXXXXXXXXXXXXX"
}

corpus_drift is the net change in result count. replay_trace_id is the ID of the new query_events row written for this replay run — inspect it with vestige-trace-show <replay_trace_id>.