Dek1milliontoken
Community@jonusnattapong · Thailand
Dek1milliontoken provides a devkit of code review, QA testing, CI repair, documentation, and multi-agent coordination skills for software engineering teams.
Agent Skills by Dek1milliontoken
Showing 38 vetted skills indexed across 3 GitHub repositories.
JonusNattapongagenttalk-consult
Send draft answers to a named peer agent for structured review.
JonusNattapongagenttalk-listen
Listen continuously for AgentTalk bus messages with identity management.
JonusNattapongagenttalk-lead
Coordinate task allocation and progress tracking among multi-agent teams in AgentTalk.
JonusNattapongagenttalk-handoff
Send structured review requests and wait for responses via AgentTalk.
JonusNattapongagenttalk-send
Send point-to-point or broadcast messages within the AgentTalk communication bus.
JonusNattapongagenttalk-propose
Send structured proposals and manage counter-proposals between agents via AgentTalk.
JonusNattapongagenttalk-sk-loop
Manage persistent task implementation and review loops for spec-kitty missions via agenttalk.
JonusNattapongtester-qa
Automate test execution and triage with evidence reporting.
JonusNattapongtest-security
Produce executable security tests and abuse-case evidence for a touched code surface.
JonusNattapongreview-docs
Verify documentation claims against current code and run documented examples.
JonusNattapongreview-code
Review code diffs for correctness, security, and maintainability with severity-tagged findings.
JonusNattapongreview-failure-injection
Evaluate code changes for error-handling and robustness across failure conditions.
JonusNattapongreview-release-readiness
Validate CI results, versioning, artifacts, permissions, documentation, and gate status for release readiness.
JonusNattapongtest-performance
Produces executed benchmark evidence for performance budgets, baselines, and suspected regressions.
JonusNattapongcraft-code
Write minimal, readable code changes matching the surrounding codebase for features, fixes, and refactors.
JonusNattapongqa-strategy
Plan test levels, review lenses, and close evidence for a code change.
JonusNattapongtest-docs
Executes documentation examples, doctests, and link checks to produce QA evidence.
JonusNattapongfix-ci
Diagnose failing CI checks and apply the smallest verified fix.
JonusNattapong_shared
Provides shared evidence profiles and skill routing references for the devkit skill suite.
JonusNattapongrefactor-code
Restructure production code while preserving observable behavior and proving preservation with tests.
JonusNattapongwrite-docs
Write and update user-facing documentation using Diataxis modes with runnable examples.
JonusNattapongassurance-scan
Automate codebase assurance scans and generate normalized JSON evidence.
JonusNattapongsystem-review-protocol
Orchestrate system reviews and aggregate evidence for HOLD/GO decisions.
JonusNattapongtest-integration
Validate behavior across real integration boundaries with isolated temp roots and deterministic commands.
Frequently Asked Questions About Dek1milliontoken
FAQPage SchemaWhat tasks can I accomplish with Dek1milliontoken's skills?▼
You can review code diffs with priority-tagged findings, write deterministic behavior-focused tests, run security and performance test evidence, diagnose and fix failing CI checks, refactor code while preserving behavior, write Diataxis-mode documentation, and coordinate multi-agent task allocation over the AgentTalk bus.
Who should use these devkit skills?▼
Software engineers, QA leads, and release managers working on code changes benefit most. The skills target practitioners who need structured review evidence, QA strategy planning, release-readiness gating, and documentation verification rather than general-purpose coding assistance.
How do the review and testing skills work in practice?▼
Each skill is scoped to a concrete trigger: review-code inspects a specific diff, test-coverage writes failing-first regression tests, fix-ci reads failing command logs before proposing fixes, and test-security produces threat-model-backed executable checks. Skills emit normalized evidence profiles like qa-result or review-result.
What prerequisites or dependencies do these skills require?▼
Skills operate on existing repositories with concrete artifacts: a code diff for review, failing CI logs for fix-ci, a defined workload and baseline for test-performance, and Python environments for doctest execution in test-docs. GPU skills require NVIDIA cuDF or cuML installations.
Do the skills cover anything beyond code quality?▼
Yes. The registry includes arXiv paper search, LangGraph documentation retrieval, web research delegation, matplotlib and seaborn chart generation, GPU-accelerated PDF extraction, SQL query generation with schema exploration, and SEO-optimized blog and social media content creation.