← Back to Skills Marketplace
Post-Development Verification
by
LeoNardo-LB
· GitHub ↗
· v1.0.1
· MIT-0
105
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install post-dev-verification
Description
Post-development full-stack verification skill. Automatically triggered after Agent completes a development task. Executes production-level validation (unit...
Usage Guidance
This skill is consistent with a test/verification tool, but it performs powerful, state-changing operations (start services, run DB migrations, seed/delete data, network calls). Before using it: (1) run Phase 0 and carefully review the Environment Report it produces; (2) ensure the agent runs in an isolated test/sandbox with no production credentials or access to production networks; (3) explicitly provide only test API keys/accounts and remove any production env vars from the runtime; (4) require an explicit human approval step before allowing Phase 2 (execution) to proceed; and (5) consider running the verification in a disposable CI environment or ephemeral container to limit blast radius. If you cannot guarantee sandboxing and credential separation, avoid running the execution phases.
Capability Analysis
Type: OpenClaw Skill
Name: post-dev-verification
Version: 1.0.1
The skill bundle defines a highly autonomous 'Post-Development Verification' agent that performs environment analysis, test generation, and an automated 'fix loop' involving code modification and service execution. While the instructions in SKILL.md are strictly aligned with software quality assurance and include safety guidelines for sandboxing, the skill requires broad, high-risk capabilities including read_project_files, run_tests, start_stop_services, and network_access. Its reliance on executing auto-detected project commands and performing multi-round code fixes constitutes a significant risk profile typical of autonomous agents with shell access, even though no malicious intent or data exfiltration was detected.
Capability Assessment
Purpose & Capability
Name/description (post-development full-stack verification) align with the SKILL.md which explains test design, environment checks, starting services, running tests and E2E validation. Declared capabilities (read_project_files, run_tests, start_stop_services, network_access) are expected for this purpose.
Instruction Scope
Instructions explicitly instruct the agent to start/stop services, run DB migrations, seed and delete test data, and make real network requests. That behavior matches the stated purpose (real-execution-first), but it grants the agent broad, potentially destructive discretion if Phase 0 review / user gating are skipped. The SKILL.md recommends Phase 0 and sandboxing, but does not enforce an explicit confirmation step before execution phases.
Install Mechanism
No install spec and no code files — instruction-only skill. This minimizes filesystem/install risk (nothing is downloaded or executed by an installer).
Credentials
requires.env is empty in registry metadata, yet the instructions and metadata repeatedly reference using test accounts, test API keys, and environment-variable-sourced tokens. Not requiring any specific env vars is coherent (they are optional), but it means the agent may try to read whatever environment variables are present; callers should explicitly provide test-only credentials and remove production credentials from the agent environment.
Persistence & Privilege
always:false and default model invocation settings (agent can invoke autonomously) are appropriate. The skill does not ask to modify other skills or system-wide settings. Its actions are limited to the project/test environment described in SKILL.md.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install post-dev-verification - After installation, invoke the skill by name or use
/post-dev-verification - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
# post-dev-verification v1.0.1
Minor update focused on safety and environment specification.
- Added explicit safety requirements, covering test isolation, test credentials, scoping of side effects, and clear user control over realism level.
- Declared metadata fields listing recommended tools, required agent capabilities, and detailed safety instructions.
- Updated documentation to emphasize running Phase 0 (environment analysis) before any execution.
- No changes to core workflow or test design methodology.
v1.0.0
Initial release of post-dev-verification skill: automated, production-grade post-development testing.
- Introduces automated, full-stack verification triggered after development completion.
- Executes unit, integration, and E2E tests prioritizing real-execution over mocks.
- Adapts realism level (L0–L3) based on user intent and environment.
- Provides structured, multi-phase workflow: environment awareness, test design, execution, feedback loop, and validation.
- Covers comprehensive test taxonomy (MFT/INV/DIR), quality metrics, and detailed anti-pattern scans.
Metadata
Frequently Asked Questions
What is Post-Development Verification?
Post-development full-stack verification skill. Automatically triggered after Agent completes a development task. Executes production-level validation (unit... It is an AI Agent Skill for Claude Code / OpenClaw, with 105 downloads so far.
How do I install Post-Development Verification?
Run "/install post-dev-verification" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Post-Development Verification free?
Yes, Post-Development Verification is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Post-Development Verification support?
Post-Development Verification is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Post-Development Verification?
It is built and maintained by LeoNardo-LB (@leonardo-lb); the current version is v1.0.1.
More Skills