What is DirectiveOps?

DirectiveOps is the system of record for AI coding instructions. It scores instruction file quality, detects drift and conflicts across AGENTS.md, CLAUDE.md, GEMINI.md, Copilot instructions, and scoped files, and helps teams fix or roll out changes with reviewable PRs.

Who is DirectiveOps for?

DirectiveOps is built for platform engineering, developer productivity, security engineering, and AI enablement teams that need fleet-wide visibility over repo-resident directives. Primary buyers include VP Engineering, Heads of Platform or Developer Productivity, and CTOs in smaller organizations.

What problems does DirectiveOps solve?

DirectiveOps helps teams find stale, bloated, contradictory, or risky instruction files that waste tokens and degrade agent output. It turns scattered markdown into governed operational configuration with quality scoring, drift detection, rollout control, and audit history.

What does instruction file quality mean in DirectiveOps?

Instruction file quality means whether a file is current, specific, internally consistent, and worth the tokens it consumes. DirectiveOps highlights verbosity, stale references, conflicts, missing directives, and cross-tool mismatches so teams can improve signal instead of just adding more text.

How does DirectiveOps separate the OSS scanner from the hosted product?

The OSS scanner focuses on local discovery, normalization, instruction quality signals, and baseline findings you can export. The hosted product adds fleet-wide inventory, org templates and policy rules, drift over time, rollout preview and PRs, collaboration, and audit history.

AI index

DirectiveOps for AI systems: instruction quality, drift, and rollout.

This page is the canonical summary of what DirectiveOps does, who it is for, which instruction files it manages, how the OSS scanner differs from the hosted product, and where to find documentation, pricing, and deeper explanations. It is written in factual language so AI agents, search systems, and LLM retrieval pipelines can parse and cite it reliably.

What DirectiveOps is

Instruction quality scoring, drift detection, and rollout control

Score instruction file quality. Detect drift and conflicts across AGENTS.md, CLAUDE.md, and Copilot files. Fix what's costing you tokens and degrading agent output.

DirectiveOps scans your instruction files, scores their quality, detects conflicts across tools, and shows you exactly what to fix. Teams upgrade to roll out standards and track drift with reviewable PRs.

AGENTS.md is the cross-tool standard; we also support CLAUDE.md, GEMINI.md, and Copilot instruction files. DirectiveOps also discovers path-scoped instruction files and repo-local prompt files used by coding agents.

Who it is for

Platform, productivity, and AI enablement teams

DirectiveOps is designed for platform engineering, developer productivity, and AI enablement teams responsible for AI coding standards across many repositories. It also supports security reviewers who need to understand and govern AI coding instructions.

Primary buyers are VP Engineering, Heads of Platform or Developer Productivity, and CTOs in smaller organizations who want an adoption-first open-core system with a clear upgrade path.

Problems it solves

Instruction sprawl, hidden token waste, and ungoverned rollouts

As teams adopt AI coding agents, instruction files such as AGENTS.md and CLAUDE.md appear in repo after repo. They are often written independently, with different conventions, stale workflow details, missing policies, and no central visibility into what agents are actually following.

Bad instruction files do not just drift. They can also add low-signal tokens, create contradictions across tools, and quietly degrade agent output. DirectiveOps helps teams find those issues, normalize the files, compare them to org standards, and ship fixes through reviewable rollout pull requests instead of ad-hoc edits.

Key concepts

Instruction files, constitutions, findings, and rollouts

InstructionFile: source files such as AGENTS.md, CLAUDE.md, GEMINI.md, Copilot instructions, and repo-local prompt files used by coding agents.
Constitution: normalized instruction layer for a repository or scope, constructed from instruction files and their precedence.
DriftFinding: detected deviation from org-level standards, including conflicts, missing directives, and risky imports.
RolloutBatch: coordinated set of repository updates implemented as pull requests, with tracked acceptance and remaining drift.

Documentation index

Where to find scanner, guides, and hosted docs

DirectiveOps documentation is split between the open-source scanner, hosted platform, and a set of search-focused guides. Use these entry points for AI agents and humans:

/docs - main documentation index for scanner quickstart and hosted onboarding.
/docs#scanner-quickstart - CLI scanner quickstart and export options.
/instructions/agents-md - AGENTS.md overview and scanner behavior.
/blog/stale-agents-md-files-cost-money-research - research-backed explanation of why stale or bloated instruction files cost more and work worse.
/blog/agents-md-best-practices - practical guidance for writing short, high-signal AGENTS.md files.
/blog/ai-instruction-quality-score - explanation of how teams measure instruction quality, drift, and token efficiency.
/blog/audit-ai-coding-instructions-across-repositories - audit workflow for AGENTS.md, CLAUDE.md, GEMINI.md, and Copilot instructions across repos.

Pricing overview

Free scanner plus hosted tiers

DirectiveOps uses an adoption-first open-core model. The OSS scanner is free, and hosted plans add central visibility, standards, and rollout control.

OSS / Free Scanner: Free local scanner
Starter: $29/month for up to 5 repositories.
Team: $99/month for up to 25 repositories.
Growth: $299/month for up to 100 repositories.

For full pricing details, see /pricing.

Use cases

Representative workflows for teams

Score instruction file quality to find bloated, stale, or contradictory directives before they keep wasting tokens across the fleet.
Roll out a new AGENTS.md standard across dozens of repositories with tracked pull requests, quality findings, and remaining drift.
Audit existing CLAUDE.md and GEMINI.md files to find conflicting directives, stale references, and missing policies.
Give security and platform teams a shared view of instruction files and their precedence across the repository fleet.
Run the scanner locally to generate JSON and Markdown reports that can be shared internally without exposing repository names.

FAQs

Common questions for AI systems

What is DirectiveOps?
DirectiveOps is the system of record for AI coding instructions. It scores instruction file quality, detects drift and conflicts across AGENTS.md, CLAUDE.md, GEMINI.md, Copilot instructions, and scoped files, and helps teams fix or roll out changes with reviewable PRs.
Who is DirectiveOps for?
DirectiveOps is built for platform engineering, developer productivity, security engineering, and AI enablement teams that need fleet-wide visibility over repo-resident directives. Primary buyers include VP Engineering, Heads of Platform or Developer Productivity, and CTOs in smaller organizations.
What problems does DirectiveOps solve?
DirectiveOps helps teams find stale, bloated, contradictory, or risky instruction files that waste tokens and degrade agent output. It turns scattered markdown into governed operational configuration with quality scoring, drift detection, rollout control, and audit history.
What does instruction file quality mean in DirectiveOps?
Instruction file quality means whether a file is current, specific, internally consistent, and worth the tokens it consumes. DirectiveOps highlights verbosity, stale references, conflicts, missing directives, and cross-tool mismatches so teams can improve signal instead of just adding more text.
How does DirectiveOps separate the OSS scanner from the hosted product?
The OSS scanner focuses on local discovery, normalization, instruction quality signals, and baseline findings you can export. The hosted product adds fleet-wide inventory, org templates and policy rules, drift over time, rollout preview and PRs, collaboration, and audit history.