VS Code Extensions

Tags

#guardrails

Glyphmap AI Guardrails

Glyphmap AI Guardrails

glyphmap-vscode

Stop AI coding agents from hallucinating files, drifting off-task, and making unsafe project changes.

Other #codex
8 5.0
LLM Watch

LLM Watch

yash-c71e123f-dc39-4c65-82bf-a7e6c9278f10

Shift-left observability and safety tooling for LLM application development.

Linters #ai
7
T

TruthSeek Eval Harness

UnifyDynamics

Configure a target model, run the TruthSeek eval suite, and view the scorecard from inside VS Code.

Testing #llm
7
rein - guardrails for AI-written code

rein - guardrails for AI-written code

sametatas

Checks the code your AI assistant writes: leaked secrets, unsafe calls, and imports that do not resolve, shown inline before you commit or run. Deterministic, no LLM. Powered by the rein engine.

Linters #python
7 5.0
Progressive Skills Editor

Progressive Skills Editor

gregoriomomm

Layered explorer for progressive-disclosure agents on watsonx Orchestrate (ADR-0015 taxonomy): AGENT (dispatcher EXPANDS to the intents it serves + exposed utility tools) ยท CELL-SKILLS (the cells, grouped by the skill-function they load) ยท SKILL-FUNCTIONS (the executables, the intent->function merge

Visualization #yaml
5
DiffGate โ€” Review Gate for AI Code

DiffGate โ€” Review Gate for AI Code

srbsa

A deterministic guardrail for AI-edited code, inline on the lines you actually changed (๐ŸŸข merge / ๐ŸŸก glance / ๐ŸŸ  verify). 0 false blocks; the same verdict as the DiffGate CLI and MCP server.

Programming Languages Linters #coding agent
4
ThumbGate

ThumbGate

igorganapolsky

Deterministic pre-action checks, feedback capture, and MCP guardrails for AI coding agents in VS Code-compatible IDEs.

Linters Machine Learning #testing
4
Velocity - AI Acceleration Layer

Velocity - AI Acceleration Layer

SuryaManne

An AI acceleration layer for Coding Assistants. Project intelligence, context layer, agent roles, workflow skills, runtime guardrails, and an SDLC pipeline โ€” version-controlled in your repo.

AI #productivity
4
Claude Code Warden

Claude Code Warden

NovisLabsIsrael

Token-usage insight and safety prompts for Claude Code: a live status-bar meter, context-window gauge, per-session budgets, and Allow/Deny prompts for risky commands.

Machine Learning Visualization #ai
3
Trinetra โ€” Architectural Guardrail

Trinetra โ€” Architectural Guardrail

utsargadhakal

Local architectural guardrails with explainable evidence for TypeScript, JavaScript, and Python. IDE sidebar, headless policy engine, and pre-commit hook โ€” no external APIs.

Linters #python
3
Prompt Compass

Prompt Compass

mpcfintech

Prompt router & firewall: keeps simple prompts on your local model, sends complex ones to the cloud, and blocks PII / jailbreaks. Uses the hosted Compass API.

Machine Learning #ai
2 5.0
aglang

aglang

Pantelispanka

Language support for aglang (.ag) โ€” architecture guardrails enforced via Z3 SMT solving

Programming Languages Linters #snippet
1