GridTRX

PassAudited by ClawScan on May 10, 2026.

Overview

GridTRX appears to be a purpose-aligned local accounting tool, but it can modify financial books and uses user-installed Python dependencies, so users should review actions before allowing writes.

Install only in a controlled workspace, keep backups of each books.db file, and have the agent ask before deleting transactions, changing lock dates, importing large batches, or running year-end rollover. The reviewed artifacts do not show hidden cloud use or exfiltration.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

An agent could change the user's accounting records inside the selected local books file.

Why it was flagged

The skill exposes write operations that can create, import, delete, and roll forward accounting records. This is expected for an accounting suite, but these are high-impact business data mutations.

Skill content
`post_transaction(...)` | Post a simple 2-line entry ... `delete_transaction(...)` | Delete a transaction ... `import_csv(...)` ... `year_end(...)` | Year-end rollover
Recommendation

Use a dedicated GRIDTRX_WORKSPACE, keep backups of books.db, and require explicit confirmation before deletes, re-imports, lock-date changes, or year-end rollovers.

What this means

A bad rule could cause future financial transactions to be categorized incorrectly until corrected.

Why it was flagged

Import rules are persistent local categorization state reused for future bank imports. Incorrect or overly broad rules could repeatedly misclassify future transactions.

Skill content
Import rules are matched by keyword (case-insensitive), highest priority wins.
Recommendation

Review import rules before bulk imports and inspect suspense/ledger results after adding or changing rules.

What this means

Different dependency versions could behave differently or introduce compatibility/security issues over time.

Why it was flagged

The optional dependencies are installed by the user and are purpose-aligned, but they are not pinned to exact versions, which can reduce reproducibility.

Skill content
flask>=2.0        # Browser UI (run.py / app.py)
mcp>=1.0          # MCP server (mcp_server.py)
Recommendation

Install dependencies in a virtual environment and consider pinning reviewed versions for production accounting use.

What this means

The browser UI may fail to start or rely on code not included in the reviewed artifact set.

Why it was flagged

The browser launcher imports app.py, but app.py is not listed in the provided file manifest. This is a packaging/provenance gap for the optional browser UI.

Skill content
from app import main as run_app
Recommendation

Use the CLI/MCP paths unless the missing browser UI source is supplied and reviewed.