0.1.9

Dependencies tree, uninstall, Chaintrap finding panel, AI-Coding Security listing. Does not block npm/pip.

0.1.6

Disable skill/rule heuristic findings (false positives). Inventory still records skills/rules; only npm/PyPI and MCP-inferred packages are flagged.

v0.1.3 — badge + open location

Highlights

• Activity-bar badge for unacknowledged high/critical findings
• Open location first; advisory is a separate button
• OSV summary projected into dependency alerts
• Expanded AI skill heuristics (Cisco-correlated, local-only)

Install: code --install-extension chaintrap-agent-shield-0.1.3.vsix

v0.1.0 — Chaintrap Agent Shield

First dogfood build of Chaintrap Agent Shield (publisher pri17m).

• Phase A baseline scan of workspace + user MCP/skills/rules
• Phase B incremental watch with OSV querybatch + known-bad fallback
• Critical findings require acknowledgment
• Agent Activity tree (baseline vs delta)

Install: code --install-extension chaintrap-agent-shield-0.1.0.vsix

Marketplace listing is blocked until a VSCE PAT with Marketplace publish scope is available for publisher pri17m.