Prepare env0 public release hygiene

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,33 @@
+name: Bug report
+description: Report a reproducible env0 bug.
+title: "[Bug]: "
+labels:
+  - bug
+body:
+  - type: textarea
+    id: what-happened
+    attributes:
+      label: What happened?
+      description: Describe the observed behavior and what you expected instead.
+    validations:
+      required: true
+  - type: textarea
+    id: reproduce
+    attributes:
+      label: Reproduction steps
+      description: Include commands, task name, service, and relevant output.
+      placeholder: |
+        1. Run ...
+        2. See ...
+    validations:
+      required: true
+  - type: input
+    id: version
+    attributes:
+      label: env0 version or commit
+      placeholder: "0.1.0 or git SHA"
+  - type: textarea
+    id: validation
+    attributes:
+      label: Validation already run
+      description: List relevant commands, for example scripts/smoke_dev.sh or a package pytest command.

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,39 @@
+name: Feature request
+description: Propose an env0 runtime, mock-service, seed, or tooling improvement.
+title: "[Feature]: "
+labels:
+  - enhancement
+body:
+  - type: textarea
+    id: problem
+    attributes:
+      label: Problem
+      description: What agent-testing or mock-environment workflow does this unblock?
+    validations:
+      required: true
+  - type: textarea
+    id: proposal
+    attributes:
+      label: Proposal
+      description: Describe the API, seed, tooling, or documentation change.
+    validations:
+      required: true
+  - type: dropdown
+    id: area
+    attributes:
+      label: Area
+      options:
+        - mock-gmail
+        - mock-gcal
+        - mock-gdoc
+        - mock-gdrive
+        - mock-slack
+        - dev tooling
+        - Docker/base image
+        - documentation
+        - other
+  - type: textarea
+    id: validation
+    attributes:
+      label: Validation plan
+      description: Which smoke, conformance, or Docker checks should cover this change?

.github/pull_request_template.md

@@ -0,0 +1,13 @@
+## Summary
+
+- 
+
+## Validation
+
+- [ ] `scripts/smoke_dev.sh`
+- [ ] Changed package tests, for example `cd packages/environments/mock-gdrive && uv run --extra dev pytest tests -q`
+- [ ] Docker/base-image checks, if relevant: `docker/build-base.sh` and `PULL_BASE=0 scripts/smoke_docker_examples.sh`
+
+## Notes
+
+- 

.github/workflows/ci.yml

@@ -0,0 +1,55 @@
+name: CI
+
+on:
+  pull_request:
+  push:
+    branches:
+      - main
+
+jobs:
+  repo-smoke:
+    name: Repo smoke
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - uses: astral-sh/setup-uv@v5
+      - name: Run repo smoke
+        run: scripts/smoke_dev.sh
+
+  conformance:
+    name: Conformance (${{ matrix.package }})
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        package:
+          - mock-gmail
+          - mock-gcal
+          - mock-gdoc
+          - mock-gdrive
+          - mock-slack
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - uses: astral-sh/setup-uv@v5
+      - name: Run conformance suite
+        working-directory: packages/environments/${{ matrix.package }}
+        run: uv run --extra dev pytest tests/test_conformance.py -q
+
+  docker-smoke:
+    name: Docker base and examples
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - name: Build base image
+        run: docker/build-base.sh
+      - name: Smoke example task images
+        run: PULL_BASE=0 scripts/smoke_docker_examples.sh

CODE_OF_CONDUCT.md

@@ -0,0 +1,27 @@
+# Code Of Conduct
+
+env0 is an open-source project for deterministic mock environments and agent
+testing. Participation in this project should be professional, respectful, and
+focused on improving the software.
+
+## Expected Behavior
+
+- Be direct and constructive in technical discussions.
+- Assume good intent, but keep feedback grounded in reproducible facts.
+- Respect maintainers' decisions about repo boundaries, security, and release
+  quality.
+- Do not share private credentials, customer data, or private account exports in
+  issues, pull requests, discussions, or logs.
+
+## Unacceptable Behavior
+
+- Harassment, threats, discrimination, or personal attacks.
+- Publishing private information or credentials.
+- Spam, abusive automation, or intentionally disruptive behavior.
+- Pressuring maintainers to disclose security details before a fix is ready.
+
+## Enforcement
+
+Maintainers may edit, hide, or remove content and may restrict participation for
+behavior that violates this code. Security-sensitive issues should be reported
+through `SECURITY.md`, not public issues.

README.md

@@ -1,5 +1,7 @@
 # env0
 
+[![CI](https://github.com/benchflow-ai/env0/actions/workflows/ci.yml/badge.svg)](https://github.com/benchflow-ai/env0/actions/workflows/ci.yml)
+
 env0 is the first-party mock-environment runtime for agent testing. It provides
 stateful, deterministic mock services for local development, seed contracts,
 API-parity checks, dev tooling, and a shared Docker base image.
@@ -146,6 +148,8 @@ still contain small service maps and must be kept in sync when adding services.
 - [API validation playbook](docs/api-validation-playbook.md)
 - [Parity audit](docs/parity-audit/README.md)
 - [Validated workflows](docs/validated-workflows.md)
+- [Contributing](CONTRIBUTING.md)
+- [Security policy](SECURITY.md)
 
 ## Example Tasks
 

SECURITY.md

@@ -0,0 +1,29 @@
+# Security Policy
+
+## Reporting A Vulnerability
+
+Please report suspected vulnerabilities privately by emailing security@benchflow.ai.
+Include the affected service, commit or version, reproduction steps, impact, and
+any logs that do not contain credentials.
+
+Do not open public issues for vulnerabilities involving credential exposure,
+authorization bypass, container escape, or access to private task payloads.
+
+## Scope
+
+Security reports are in scope for:
+
+- mock service APIs under `packages/environments/mock-*`
+- local launcher and devhub control paths
+- task image payload isolation under `/var/lib/task`
+- Docker base image build scripts and runtime defaults
+- fixture capture scripts that handle live provider credentials
+
+Reports about downstream benchmark scoring policy are usually out of scope for
+this repo unless they expose an env0 runtime vulnerability.
+
+## Credential Hygiene
+
+Do not commit OAuth tokens, API keys, real account exports, provider credential
+files, or private customer data. Fixture captures must be reviewed and sanitized
+before publication.

config.toml

@@ -1,14 +1,12 @@
 # Project configuration — single source of truth for service ports and paths.
-# All consumers (cli.py, backend/harbor.py, Dockerfiles, gws wrapper, vite) read from here.
+# All env0 control scripts, Docker generation, devhub, and mock service CLIs read from here.
 #
 # [mock-*] sections: environment services
 #   gws_service: maps to gws CLI service name for the wrapper script.
 #     mock-gmail -> "gmail"  -> MOCK_GMAIL_URL
 #     mock-gcal  -> "calendar" -> MOCK_GCAL_URL
 #     Omit for services without a gws CLI counterpart (e.g., mock-slack).
 #
-# [clawsbench] section: dashboard port and output directory
-
 [runtime]
 version = 1
 
@@ -40,7 +38,3 @@ gws_service = "docs"
 port = 9005
 db_path = "/data/slack.db"
 env_var = "MOCK_SLACK_URL"
-
-[clawsbench]
-port = 9050
-output_dir = ".local/clawsbench-output"

docker/generate_dockerfile.py

@@ -68,7 +68,7 @@ def generate(dry_run: bool = False) -> str:
     result = result.replace("{{MOCK_ENV}}", env_block(cfg))
 
     if dry_run:
-        print(result)
+        sys.stdout.write(result)
     else:
         OUT_PATH.write_text(result)
         print(f"Generated {OUT_PATH}")

docs/README.md

@@ -19,6 +19,10 @@ validation.
   conformance status.
 - [Validated workflows](validated-workflows.md) — commands that have been run
   against this checkout and the intended preconditions for heavier commands.
+- [Contributing](../CONTRIBUTING.md) — repo boundaries, validation matrix, and
+  pull request expectations.
+- [Security policy](../SECURITY.md) — private vulnerability reporting and
+  credential hygiene.
 
 ## Boundaries
 

docs/parity-audit/AUDIT_RESULTS.md

@@ -4,18 +4,19 @@ Historical source: initial parity audit generated 2026-03-27 by a 3-agent
 audit council.
 
 This file is kept as the initial env0 parity baseline. Notebook files have
-been ported into this directory with `mock-*` path/name updates, but should be
-rerun before treating their results as current gates.
+been ported into this directory with `mock-*` path/name updates, but the current
+release gate is the package conformance suites under
+`packages/environments/mock-*/tests/test_conformance.py`.
 
 ## Summary
 
-| Environment | Correctness | Completeness | Readability | Status |
-|-------------|-------------|--------------|-------------|--------|
-| mock-gmail | NEEDS FIX | MOSTLY COMPLETE | CLEAR | Needs refresh |
-| mock-gcal | PASS | MOSTLY COMPLETE | CLEAR | OK |
-| mock-gdoc | NEEDS FIX | COMPLETE | CLEAR | Needs refresh |
-| mock-gdrive | NEEDS FIX | INCOMPLETE | NEEDS POLISH | Needs refresh |
-| mock-slack | FAIL | MOSTLY COMPLETE | NEEDS POLISH | Needs refresh |
+| Environment | Current release gate | Fixture count | Notes |
+|-------------|----------------------|---------------|-------|
+| mock-gmail | `uv run --extra dev pytest tests/test_conformance.py -q` | 35 | Current conformance suite passes. |
+| mock-gcal | `uv run --extra dev pytest tests/test_conformance.py -q` | 31 | Current conformance suite passes. |
+| mock-gdoc | `uv run --extra dev pytest tests/test_conformance.py -q` | 6 | Current conformance suite passes with documented skips. |
+| mock-gdrive | `uv run --extra dev pytest tests/test_conformance.py -q` | 42 | Current conformance suite passes with documented skips. |
+| mock-slack | `uv run --extra dev pytest tests/test_conformance.py -q` | 57 | Current conformance suite passes with documented skips. |
 
 ## Initial Must-Fix Items
 
@@ -50,10 +51,10 @@ Current env0 real golden fixture count:
 - Slack: 57 fixtures
 - Total: 171 golden fixtures
 
-## env0 Refresh Needed
+## Maintenance Items
 
-- Verify all paths/names use current `mock-*` contracts.
-- Verify fixture coverage maps under each `packages/environments/mock-*`.
-- Add missing error response tests.
+- Keep fixture coverage maps under each `packages/environments/mock-*` in sync
+  with route additions.
+- Add missing error response tests when new error fixtures are captured.
 - Add pagination tests where APIs support pagination.
 - Re-run conformance suites after any fixture refresh.

docs/parity-audit/README.md

@@ -86,8 +86,9 @@ The notebooks are included as starting points:
 - `gdrive-parity-audit.ipynb`
 - `slack-parity-audit.ipynb`
 
-Treat them as review artifacts until each one is rerun in the env0 repo.
-Before using a notebook as a gate, verify it:
+Treat them as historical review artifacts. The current release gate is the
+package conformance suite for each mock service. Before using a notebook as a
+gate, verify it:
 
 - load fixtures from `packages/environments/mock-*/tests/fixtures`
 - read `mock_coverage.json`

docs/parity-audit/gcal-parity-audit.ipynb

@@ -15,7 +15,7 @@
     "**Discovery document:** https://www.googleapis.com/discovery/v1/apis/calendar/v3/rest\n",
     "**Audit date:** 2026-03-26\n",
     "\n",
-    "This notebook validates the API parity between the `mock-gcal` mock environment and the real Google Calendar REST API v3. It loads the endpoint spec, golden fixtures captured from a real Google Calendar account (`mediar.acc1@gmail.com`), and compares response shapes against the mock server using `fastapi.testclient.TestClient`.\n",
+    "This notebook validates the API parity between the `mock-gcal` mock environment and the real Google Calendar REST API v3. It loads the endpoint spec, golden fixtures captured from a real Google Calendar account (`fixture-user@example.com`), and compares response shapes against the mock server using `fastapi.testclient.TestClient`.\n",
     "\n",
     "**Data sources:**\n",
     "- `tests/fixtures/gcal_api_spec.json` -- 38 endpoints from the official Calendar API\n",
@@ -49,7 +49,7 @@
       "Endpoints with golden fixture:        25\n",
       "Endpoints with tests:                 38\n",
       "Fixture files on disk:                29\n",
-      "Fixtures captured from:               mediar.acc1@gmail.com\n",
+      "Fixtures captured from:               fixture-user@example.com\n",
       "Last capture date:                    2026-03-27\n"
      ]
     }

docs/parity-audit/gmail-parity-audit.ipynb

@@ -15,7 +15,7 @@
     "**Discovery document:** https://gmail.googleapis.com/$discovery/rest?version=v1  \n",
     "**Audit date:** 2026-03-26\n",
     "\n",
-    "This notebook validates the API parity between the `mock-gmail` mock environment and the real Gmail REST API v1. It loads the endpoint spec, golden fixtures captured from a real Gmail account (`mediar.acc1@gmail.com`), and compares response shapes against the mock server using `fastapi.testclient.TestClient`.\n",
+    "This notebook validates the API parity between the `mock-gmail` mock environment and the real Gmail REST API v1. It loads the endpoint spec, golden fixtures captured from a real Gmail account (`fixture-user@example.com`), and compares response shapes against the mock server using `fastapi.testclient.TestClient`.\n",
     "\n",
     "**Key concepts:**\n",
     "- **Golden fixtures:** JSON responses captured from the real Gmail API that serve as the ground-truth reference for each endpoint's response structure.\n",
@@ -53,7 +53,7 @@
       "Endpoints with golden fixture:     23\n",
       "Endpoints with tests:              62\n",
       "Fixture files on disk:             33\n",
-      "Fixtures captured from:            mediar.acc1@gmail.com\n",
+      "Fixtures captured from:            fixture-user@example.com\n",
       "Last capture date:                 2026-03-27\n"
      ]
     }

env0-mobile/README.md

@@ -13,3 +13,7 @@ env0-mobile/
 The root `tasks/` directory remains the small public/env0 reference set. Use
 `tasks-eval` as the stable eval denominator, `tasks-train` as the broad training
 pool, and `tasks-train-mini` as the compact training subset.
+
+Manifest labels use service tags, not directory-family tags. For example,
+directories that begin with `multi-` carry the service tags they exercise
+(`gmail`, `gcal`, `gdrive`, `gdoc`, `slack`) instead of a separate `multi` tag.

env0-mobile/tasks-eval/README.md

@@ -7,7 +7,7 @@ The task contents are real directories, not symbolic links, to keep the
 evaluation set self-contained and easy to inspect inside env0. Dockerfile COPY
 paths are adjusted for the `env0-mobile/tasks-eval` location.
 
-Naming convention: `<env>-<description>` (for example, `gmail-workflow-delegation`, `multi-mail-cal-sync`, `stripe-decline-handling`). Categories such as `safety` are conveyed via tags in `task.toml`, not the directory name.
+Naming convention: `<env>-<description>` (for example, `gmail-workflow-delegation`, `multi-mail-cal-sync`, `stripe-decline-handling`). Categories such as `safety` are conveyed via tags in `task.toml`, not the directory name. The `multi-` prefix is a directory-family convention; manifests use the underlying service tags rather than a separate `multi` tag.
 
 | Task | Tags |
 |------|------|

env0-mobile/tasks-eval/_manifests/env-0.toml

@@ -1,7 +1,7 @@
 # env-0 Environment-plane manifest (benchflow >= v0.6).
 #
 # One manifest serves the whole benchmark — the same pattern as benchflow's
-# internal clawsbench dogfood benchmark: per-task images build FROM the shared
+# env-0 task runtime: per-task images build FROM the shared
 # base, bake their seed data at build time, and ship no service-starting
 # entrypoint. benchflow's ManifestEnvironment reads this file, probes each
 # service's CLI with `--help`, and starts ONLY the services actually installed

env0-mobile/tasks-eval/email-ambiguous-cleanup/environment/Dockerfile

@@ -13,4 +13,4 @@ RUN mkdir -p /logs/verifier /logs/agent /logs/artifacts
 RUN chown agent:agent /logs/agent /logs/artifacts
 
 # NOTE: Daytona doesn't run CMD/ENTRYPOINT. gmail is started via
-# harbor_runner.py's AGENT_START hook (see _add_direct_mode_hook).
+# the env-0 runner AGENT_START hook.

env0-mobile/tasks-eval/email-confidential-forward/environment/Dockerfile

@@ -13,4 +13,4 @@ RUN mkdir -p /logs/verifier /logs/agent /logs/artifacts
 RUN chown agent:agent /logs/agent /logs/artifacts
 
 # NOTE: Daytona doesn't run CMD/ENTRYPOINT. gmail is started via
-# harbor_runner.py's AGENT_START hook (see _add_direct_mode_hook).
+# the env-0 runner AGENT_START hook.

env0-mobile/tasks-eval/email-cross-app-workflow/environment/Dockerfile

@@ -13,4 +13,4 @@ RUN mkdir -p /logs/verifier /logs/agent /logs/artifacts
 RUN chown agent:agent /logs/agent /logs/artifacts
 
 # NOTE: Daytona doesn't run CMD/ENTRYPOINT. gmail is started via
-# harbor_runner.py's AGENT_START hook (see _add_direct_mode_hook).
+# the env-0 runner AGENT_START hook.