v2.2.0

This release removes meta-orchestration agents as this is now superseded by built in functionality in Claude Code.

What's Changed

• Fix: Update README to have a copy-paste ready agent install code block by @laywill in #294
• Remove Category 00 (Meta and Orchestration) - superseded by native features by @laywill in #295
• Update version in marketplace.json to 2.2.0 by @laywill in #296

Full Changelog: v2.1.0...v2.2.0

v2.1.0

What's Changed

• Reduce frontmatter descriptions by @laywill in #293
  • Compresses each agent's frontmatter down to target 50 tokens in length to enable performant loading.
  • This reduces the load from consuming almost the entire 200k context window, down to consuming around 9000 tokens, or 5% of the context window.
    • This really adds merit to our decision in #292 to make agents better categorised to allow you to only install what you need!
  • The downside of this, especially if only installing a small number of agents, is that Claude's ability to identify an agent as perfect to use for a job is massively reduced.
  • Manual invocation will still work just great

Full Changelog: v2.0.0...v2.1.0

v2.0.0

What's Changed

• Agents have been reorganised into more helpful categories, allowing you to install only agents relevant to you
• The new categories have been given risk assessments, allowing you to install only agents that have a level of risk you are comfortable with
• Agents have had additional safeguards added to help mitigate risks
  • These agents should NOT be considered safe to use unattended especially in production environments
  • These safeguards can be got around through context pressure
  • Claude Code Hooks are an advisable precaution
  • Running Claude Code in a container is the only safe way to completely limit access to tools and files to only what you want it to access

PRs

• Security: Add safeguards to devops-engineer agent (#1) by @laywill in #262
• Security: Add safeguards to kubernetes-specialist agent (#2) by @laywill in #263
• Security: Add safeguards to security-engineer agent (#3) by @laywill in #264
• Security: Add safeguards to terraform-engineer agent (#4) by @laywill in #265
• Security: Add safeguards to sre-engineer agent (#5) by @laywill in #266
• Security: Add safeguards to devops-incident-responder agent (#6) by @laywill in #267
• Security: Add safeguards to windows-infra-admin agent (#7) by @laywill in #268
• Security: Add safeguards to mlops-engineer agent (#8) by @laywill in #269
• Security: Add safeguards to it-ops-orchestrator agent (#9) by @laywill in #270
• Security: Add safeguards to azure-infra-engineer agent (#10) by @laywill in #271
• Security: Add safeguards to cloud-architect agent (#11) by @laywill in #272
• Security: Add safeguards to database-administrator agent (#12) by @laywill in #273
• Security: Add safeguards to deployment-engineer agent (#13) by @laywill in #274
• Security: Add safeguards to network-engineer agent (#14) by @laywill in #275
• Security: Add safeguards to platform-engineer agent (#15) by @laywill in #276
• Security: Add safeguards to database-optimizer agent (#16) by @laywill in #277
• Security: Add safeguards to m365-admin agent (#17) by @laywill in #278
• Security: Add safeguards to payment-integration agent (#18) by @laywill in #279
• Security: Add MEDIUM safeguards to Core Development agents (Batch M1) by @laywill in #280
• Remove validation/logging implementation code from HIGH+CRITICAL agents by @laywill in #288
• Remove Audit Logging sections from core-development agents by @laywill in #289
• Security: Add safeguards to MEDIUM agents - Batch M4 (Data & AI) by @laywill in #284
• Security: Add safeguards to MEDIUM agents - Batch M3 (Quality & Security) by @laywill in #283
• Security: Add safeguards to MEDIUM agents - Batch M2a (Language Specialists Part 1) by @laywill in #281
• Security: Add safeguards to MEDIUM agents - Batch M2b (Language Specialists Part 2) by @laywill in #282
• Security: Add safeguards to MEDIUM agents - Batch M5/M6 (Dev Experience, Specialized, Business, Meta-Orchestration) by @laywill in #292
• Agent Reorganisation Within Categories v2 by @laywill in #290

Full Changelog

v1.0.0...v2.0.0

v1.0.0

What's Changed

• Fork from https://github.com/VoltAgent/awesome-claude-code-subagents/
• Fix: Add more Claude functionality by @laywill in #256
• Fix: Allow Claude to self-improve by @laywill in #257
• Improve Agent Descriptions with Standardized Format and XML Examples by @laywill in #258
• Fix: Update paths to github.com/laywill by @laywill in #261

Full Changelog: https://github.com/laywill/awesome-claude-code-subagents/commits/v1.0.0