Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .codex-plugin/plugin.json
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
{
"name": "reflex",
"version": "0.17.1",
"version": "0.18.0",
"description": "Prod-safe AI coding agents for infra teams: a pre-execution hook that judges each shell command by where it points and what it will change, then runs it, asks a human, or blocks it.",
"author": {
"name": "Ciprian Ursu",
Expand Down
5 changes: 4 additions & 1 deletion CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,8 @@ All notable changes to Reflex are documented here. The format follows

## [Unreleased]

## [0.18.0] - 2026-09-30

### Changed

- Laya engine: `reflex setup --engine laya` installs `laya[serve]==0.3.22` (was 0.3.20). Measured head to head with 0.3.20 and Jev 1.13.0 on every golden set on 2026-09-30: every number is the same, and every answer is the same except four gate cases whose state holds a per-process path (at most 0.013 of probability, no decision changed). The checkpoints stay at revision `55cf4c4`, still the latest. The GUIDE's "Measured against Jev" table is rerun on the current golden sets (199 gate cases, 47 ladder commands) and adds the MCP set, where calibrated Laya matches Jev (0 MISS, 0 over-strict, on 6 cases that reach System 1). Laya stays off by default.
Expand Down Expand Up @@ -1086,7 +1088,8 @@ Fourth review of the read-only parser (#43). Each change asks more often; none p
and 3.12).
- `LICENSE` (MIT), `SECURITY.md`, `CONTRIBUTING.md`, this changelog and issue templates.

[Unreleased]: https://github.com/ursuciprian/reflex/compare/v0.17.1...HEAD
[Unreleased]: https://github.com/ursuciprian/reflex/compare/v0.18.0...HEAD
[0.18.0]: https://github.com/ursuciprian/reflex/compare/v0.17.1...v0.18.0
[0.17.1]: https://github.com/ursuciprian/reflex/compare/v0.17.0...v0.17.1
[0.17.0]: https://github.com/ursuciprian/reflex/compare/v0.16.0...v0.17.0
[0.16.0]: https://github.com/ursuciprian/reflex/compare/v0.15.0...v0.16.0
Expand Down
6 changes: 3 additions & 3 deletions action.yml
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,7 @@ runs:
- name: Validate the team policy
shell: bash
env:
REFLEX: npx --yes @ursuciprian/reflex@0.17.1
REFLEX: npx --yes @ursuciprian/reflex@0.18.0
REFLEX_DATA_DIR: ${{ runner.temp }}/reflex
XDG_CONFIG_HOME: ${{ runner.temp }}/reflex-config
DIR: ${{ inputs.directory }}
Expand Down Expand Up @@ -80,7 +80,7 @@ runs:
id: check
shell: bash
env:
REFLEX: npx --yes @ursuciprian/reflex@0.17.1
REFLEX: npx --yes @ursuciprian/reflex@0.18.0
REFLEX_DATA_DIR: ${{ runner.temp }}/reflex
XDG_CONFIG_HOME: ${{ runner.temp }}/reflex-config
REFLEX_ENGINE: local
Expand Down Expand Up @@ -127,7 +127,7 @@ runs:
if: inputs.golden-eval == 'true'
shell: bash
env:
REFLEX: npx --yes @ursuciprian/reflex@0.17.1
REFLEX: npx --yes @ursuciprian/reflex@0.18.0
REFLEX_DATA_DIR: ${{ runner.temp }}/reflex
XDG_CONFIG_HOME: ${{ runner.temp }}/reflex-config
TYPESAFE_API_KEY: ${{ inputs.typesafe-api-key }}
Expand Down
2 changes: 1 addition & 1 deletion docs/SETUP.md
Original file line number Diff line number Diff line change
Expand Up @@ -567,7 +567,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: ursuciprian/reflex@v0.17.1 # a release tag or a commit SHA
- uses: ursuciprian/reflex@v0.18.0 # a release tag or a commit SHA
with:
commands: |
make deploy-staging
Expand Down
2 changes: 1 addition & 1 deletion package.json
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
{
"name": "@ursuciprian/reflex",
"version": "0.17.1",
"version": "0.18.0",
"description": "Prod-safe AI coding agents for infra teams: terraform AI agent guardrails (plan-aware apply gate), AWS, kubectl and Terraform workspace context, change freezes, a team policy and a SOC 2 / ISO 27001 audit export for Claude Code, Codex CLI, opencode and pi. Pre-execution risk gate and prompt injection guard with local rules and optional TypeSafe Jev",
"keywords": [
"ai-coding-agents",
Expand Down
2 changes: 1 addition & 1 deletion plugin/.claude-plugin/plugin.json
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
{
"name": "reflex",
"displayName": "Reflex",
"version": "0.17.1",
"version": "0.18.0",
"description": "Prod-safe AI coding agents for infra teams: a pre-execution hook that judges each shell command by where it points and what it will change, then runs it, asks a human, or blocks it.",
"author": {
"name": "Ciprian Ursu",
Expand Down
2 changes: 1 addition & 1 deletion plugin/package.json
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
{
"name": "@ursuciprian/reflex",
"version": "0.17.1",
"version": "0.18.0",
"private": true,
"type": "module",
"license": "MIT"
Expand Down
Loading