Catalogskills.sh · parcadei

recall

Query the memory system for relevant learnings from past sessions

Review advisedVerified hub495 installs

Savant verdict: Review advised

Security analysis or source-hub review identified patterns requiring attention.

Live evaluation

Not evaluated yet. Workspaces can request a live evaluation.

Security assessment

Scan pending.

Structure

  • 1 referenced file isn't in the packagescripts/core/recall_learnings.py. They may be binary, too large to catalog, or missing upstream.
  • No license declaredConfirm you may reuse this skill before importing it into your repository.
  • The source hub's security checks flagged this skillGen Agent Trust Hub: fail (The skill is vulnerable to command injection because it directly inserts user input into a shell command. It also lacks protection against potentially malicious instructions stored in the retrieved memory data.); Runlayer: warn (1/1 file flagged)

SKILL.md

---
name: recall
description: Query the memory system for relevant learnings from past sessions
user-invocable: false
---

# Recall - Semantic Memory Retrieval

Query the memory system for relevant learnings from past sessions.

## Usage

```
/recall <query>
```

## Examples

```
/recall hook development patterns
/recall wizard installation
/recall TypeScript errors
```

## What It Does

1. Runs semantic search against stored learnings (PostgreSQL + BGE embeddings)
2. Returns top 5 results with full content
3. Shows learning type, confidence, and session context

## Execution

When this skill is invoked, run:

```bash
cd $CLAUDE_OPC_DIR && PYTHONPATH=. uv run python scripts/core/recall_learnings.py --query "<ARGS>" --k 5
```

Where `<ARGS>` is the query provided by the user.

## Output Format

Present results as:

```
## Memory Recall: "<query>"

### 1. [TYPE] (confidence: high, id: abc123)
<full content>

### 2. [TYPE] (confidence: medium, id: def456)
<full content>
```

## Options

The user can specify options after the query:

- `--k N` - Return N results (default: 5)
- `--vector-only` - Use pure vector search (higher precision)
- `--text-only` - Use text search only (faster)

Example: `/recall hook patterns --k 10 --vector-only`