Live survey304 models · 14 providers · 23,592 auditsAudited implement · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

Memtrace Episode Replay

skill · syncable-dev/memtrace-publicNOASSERTION licenceAudited 27 Oct 2026 · source 3e7894d935d8

★ 505pushed today
!CAUTIONGrade B · 89/100 · 1 finding · 3e7894d
505repo stars · whole syncable-dev/memtrace-public

Is it safe? With care — CAUTION, grade B (89/100). We read this skill's source at commit 3e7894d935d8 and found no filesystem, network or shell access and no hard-coded secrets. The audit raised 1 finding, 0 critical or high. What it does: Replay the graph diff of one episode — a single git commit or working-tree save — to inspect what it changed: added/modified/removed symbols and edges. Use when the user asks what one commit or save changed in the…

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Memtrace Episode Replay (syncable-dev/memtrace-public)”. Retrieved 27 Oct 2026. https://modelsatlas.com/skills/syncable-dev/memtrace-episode-replay

Trust audit

Audit 0.4.1 · source read 27 Oct 2026

Trust score89/100!CAUTION B
0 critical0 high1 medium0 low0 info

CAUTION means the audit found things worth knowing before you trust this skill. Grade B is shared by 17,406 of 18,389 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesWARN1
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (1)

MEDIUMPrompt injectionreview.instruction_override · <listing:description>
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Do not use git log, git diff, Grep, or manual history reconstruction; Memtrace has episodic symbol replay.
Why
The package unconditionally instructs the agent to stop using its normal investigation tools in favor of a specific third-party workflow.
Fix
rewrite it so the instruction says plainly what it does, and asks the user before it acts

Audit 0.4.1 · source sha 3e7894d935d8 · 27 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publishersyncable-dev GitHub organizationOwner of the repository this skill is read from
  • Source foldersyncable-dev/memtrace-public/plugins/memtrace-skills/skills/memtrace-episode-replayThe exact folder we audited
  • LicenceNOASSERTIONAs the repository states it
  • Commit pinned for this verdict3e7894d935d8A new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Replay the graph diff of one episode — a single git commit or working-tree save — to inspect what it changed: added/modified/removed symbols and edges. Use when the user asks what one commit or save changed in the…

---
name: memtrace-episode-replay
description: "Replay the graph diff of one episode — a single git commit or working-tree save — to inspect what it changed: added/modified/removed symbols and edges. Use when the user asks what one commit or save changed in the graph, why code looks this way, or wants to inspect implementation attempts, reversions, past reasoning, or abandoned approaches across commits and working-tree episodes. Do not use git log or Grep for graph-level episode diffs; Memtrace replays indexed episode records. Do NOT use for module-level summaries or date-range change history — use memtrace-evolution."
---

## Overview

Replay the graph diff for **one episode** (a single git commit or working-tree save). Shows which nodes/edges were added, modified, or removed in that episode — not a multi-episode time-range narrative.

Use `get_evolution(mode: "recent")` to **find** episode IDs and timestamps, then drill into a specific episode with this tool.

## Episode selector — pick one

| Approach | Parameters |
|---|---|
| Known episode UUID | `episode_id` |
| Newest episode in repo | `repo_id` + `episode_index: 0` |
| Nth newest | `repo_id` + `episode_index: N` |

**There is no `from` / `to` / `include_working_tree` on this tool.**

## Optional filters (large commits)

| Param | Purpose |
|---|---|
| `symbol` | Scope to one symbol name |
| `file_path` | Substring filter on record paths |
| `kind` | e.g. `"Function"`, `"CALLS"` |
| `mode` | `"graph_summary"` for digest on huge commits |
| `compress` | default `true` — collapse identical-hash modifications |
| `limit` / `cursor` | Pagination per bucket (default limit 200) |

Full parameter spec for every Memtrace tool: `references/mcp-parameters.md` (bundled at the memtrace-skills plugin root).

## Steps

### 1. Find the episode to inspect

From `get_evolution`:

```json
{ "repo_id": "memdb", "from": "7d ago", "mode": "recent", "limit": 20 }
```

Each entry has `episode` metadata including `id` (use as `episode_id`) and `reference_time`.

Or use `episode_index: 0` for the newest episode without knowing the UUID.

### 2. Call `get_episode_replay`

By episode UUID:

```json
{
  "episode_id": "550e8400-e29b-41d4-a716-446655440000",
  "symbol": "execute",
  "file_path": "src/order/service.rs",
  "kind": "Function",
  "compress": true,
  "limit": 200
}
```

Newest episode, summary first on large commits:

```json
{
  "repo_id": "memdb",
  "episode_index": 0,
  "mode": "graph_summary"
}
```

### 3. Interpret the response

| Field | Meaning |
|---|---|
| `found` | `false` if episode missing — check `_note` |
| `totals` | Counts: `nodes_added/modified/removed`, `edges_*` |
| `nodes_added[]` etc. | Per-record diffs (paginated) |
| `page.next_cursor` | More records remain — pass as `cursor` |

For symbol history **across many episodes**, use `get_timeline` instead:

```json
{
  "repo_id": "memdb",
  "scope_path": "OrderService.execute",
  "file_path": "src/order/service.rs"
}
```

## Output

`get_episode_replay` returns the buckets interpreted in step 3:

```json
{
  "found": true,
  "totals": { "nodes_added": 3, "nodes_modified": 12, "nodes_removed": 1, "edges_added": 7, "edges_removed": 2 },
  "nodes_added": [ /* per-record diffs, paginated per bucket */ ],
  "nodes_modified": [ /* ... */ ],
  "edges_removed": [ /* ... */ ],
  "page": { "next_cursor": 200 }
}
```

## Common Mistakes

| Mistake | Reality |
|---------|---------|
| Passing `from` / `to` time window | Not supported — one episode per call |
| Expecting `narrative_hint` / `attempted_and_reverted` | Not in API — inspect added/modified/removed buckets |
| Unfiltered replay on 10k-symbol commits | Use `kind`, `file_path`, `symbol`, or `mode: "graph_summary"` first |
| Using this for "what changed last week?" or module summaries | Use memtrace-evolution — `get_evolution(from=..., mode: recent)` — to list episodes |

Excerpt, first 3,892 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 3e7894d

Download folder ↗
git clone https://github.com/syncable-dev/memtrace-public
git -C memtrace-public checkout 3e7894d935d8

The author documents no install command, so this uses the repository directly. Our verdict covers commit 3e7894d935d8 only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
CursorMentioned in its text, not testedSkills for Cursor · About Cursor
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. !CAUTION B 89 / 100

    first audit · source 3e7894d935d8

README badge

For maintainers · shows the latest audit

ModelsAtlas: CAUTION BMODELSATLASCAUTION · B
[![ModelsAtlas: CAUTION B](https://modelsatlas.com/skills/syncable-dev/memtrace-episode-replay/badge.svg)](https://modelsatlas.com/skills/syncable-dev/memtrace-episode-replay/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Memtrace Relationships✓SAFE

syncable-dev/memtrace-public

Map source-code relationships between symbols. Use when the user asks about callers, callees, references, imports, exports, type usages, class hierarchy, inheritance, implementations, overrides, or dependencies between…

★ 509Grade B
Memtrace Search✓SAFE

syncable-dev/memtrace-public

Find source code with Memtrace hybrid BM25+semantic search: symbols, functions, classes, types, constants, definitions, implementations, logic, or error strings inside code. Use when the user wants to find, search,…

★ 509Grade B
Memtrace Session Continuity✓SAFE

syncable-dev/memtrace-public

Catch up on everything that changed in an indexed source-code repo since the last session, using stored session anchors and Memtrace change memory. Use when the user asks to continue, catch up, resume, see what changed…

★ 509Grade B
Memtrace Index✓SAFE

syncable-dev/memtrace-public

Index a source-code repo into the Memtrace knowledge graph and poll the job to completion. Use when the user asks to index, parse, ingest, reindex, watch, or prepare a source-code repo for Memtrace analysis, when code…

★ 508Grade B
Memtrace Quality✓SAFE

syncable-dev/memtrace-public

Find dead code, complexity hotspots, and refactoring candidates in indexed source code. Use when the user asks about source-code quality, dead code, unused functions, zero callers, complexity, cyclomatic complexity,…

★ 508Grade B
Memtrace Refactoring Guide✓SAFE

syncable-dev/memtrace-public

Build a phased, risk-scored refactoring plan from Memtrace complexity, dead-code, bridge, impact analysis, and Cortex decision-memory constraints. Use when the user wants to refactor source code, reduce complexity,…

★ 508Grade B

Add this badge to your README

[![ModelsAtlas: CAUTION B](https://modelsatlas.com/skills/syncable-dev/memtrace-episode-replay/badge.svg)](https://modelsatlas.com/skills/syncable-dev/memtrace-episode-replay/)

Questions about Memtrace Episode Replay

What does the Memtrace Episode Replay skill do?
Structural memory for AI coding agents. Bi-temporal graph, MCP-native, zero LLM calls. Cursor · Claude Code · Codex · DeepSeek Harness · Hermes · VS Code · Windsurf.
Is Memtrace Episode Replay safe to install?
With care. The audit graded it B (89/100) and found 1 thing worth knowing before you trust this skill, listed below with the exact line each was found on.
What can Memtrace Episode Replay access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does Memtrace Episode Replay work with?
Its documentation mentions cursor. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (3e7894d935d8), read on 2026-10-27. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.