Writing

The ideas behind the platform.

The writing arm of QA Veritas — an exploration of AI-Native Verification Engineering.

Field notes on a future where humans and AI agents operate complex systems together, and the engineering substrate — operational memory, deterministic reasoning, observable verification, durable runbooks — that makes that collaboration trustworthy. Each piece pairs with a runnable component of the platform:

Resource Ledger · State Triage · LogLens · Intent Verify · Runbook Forge · SkillPack.

Articles

#	Title
01	Why QA Is Becoming Infrastructure Engineering
02	The Deterministic Pipeline Around a Nondeterministic Agent
03	Resource Memory: Operating Infrastructure From a Git Tree
04	Read Before Write: Safety Loops for Agents That Touch Production
05	Code-Aware Log Intelligence: Stop Reading Multi-Gigabyte Logs
06	A Failure Taxonomy for Distributed-System Test Runs
07	Progressive-Disclosure Skills: An Architecture for Reusable Agent Capability
08	Intent-Driven Testing: From Writing Tests to Specifying Verification
09	Verification as a First-Class Artifact
10	Capacity-Aware Feasibility Checks Before Any Change
11	Resumable Agent Workflows: State That Survives the Session
12	Institutional Memory That Outlives Turnover
13	The Triage Agent: How to Investigate a Failure Like an Engineer
14	Non-Prescriptive Skills: Give Agents Knowledge, Not Command Lists
15	Engineering Leverage: One Engineer, A Fleet of Agents
16	Root-Cause Analysis as a Repeatable System
17	Observability Interpretation: Turning Signals Into Decisions
18	Upgrade Automation as a Verified State Transition
19	The Incident Replay Loop: Learning From Every Failure
20	Building an AI-Native Verification Engineering Practice From Scratch

Playbooks

Step-by-step procedures, each tied to a tool in the repos above.

#	Title
01	Triage a Failed Test Run
02	Safely Change a Live Resource
03	Capacity Feasibility Check
04	Code-Aware Log Investigation
05	Generate a Runbook from History
06	Author a Progressive-Disclosure Skill
07	Resumable Workflow State
08	Verified Upgrade as a State Transition

Talks

Slide decks (Markdown / reveal-style).

#	Title
01	AI-Native Verification Engineering
02	Deterministic Pipelines Around Agents
03	Resource Memory and Safety Loops
04	From Tests to Intent and Verification

Agent designs

Reference designs for specialized agents — what each is responsible for, the tools and skillpack skills it uses, its workflow, and guardrails. The orchestration layer that sits above the repos.

Agent	Responsibility
QA	Specify intent, verify outcomes, own the investigation trail.
SRE	Keep services healthy; read-before-write changes with observable verification.
Infrastructure	Operate resources from the git-tracked ledger; capacity-aware feasibility.
Verification	Turn intent into observable checks; verified / failed / inconclusive.
Upgrade	Drive upgrades as verified, resumable state transitions.
Incident	Capture state during an incident; drive the replay loop.
Triage	Parse facts, classify, plan cheapest-first, synthesize root cause.
Knowledge	Maintain institutional memory in the repo, not in people.
Observability	Interpret signals into decisions; correlate metrics, logs, change history.

---

MIT licensed. Written by Ajay Singh.