mirror of
https://github.com/boshu2/agentops.git
synced 2026-09-14 15:08:13 +08:00
Bo
3213afcf1c
## Change Make native coding-agent execution the default AgentOps entry path with zero mandatory skills. Preserve full bundles and add repeatable `ao skills link --skill NAME` selection, validating the entire selection before writes. Align product, installation, architecture and generated command documentation. Extend the existing trial readout to separate endpoint test results, execution state and independently accepted work. Bind supplied judgments to exact content, acceptance and native evidence. Reject empty implementation subjects and require the caller's complete criterion ID set before reporting acceptance. Preserve genuine nonempty and deletion-only subjects, valid failures and missing-proof outcomes. ## Validation - Native onboarding from empty home/consumer directories produces no setup files; selective/full linking and failure boundaries are covered. - Actual RED/GREEN regressions cover empty subjects and the partial-criterion omission found by independent review. - Full Go build, vet and race/shuffle tests; affected Go lint; 88 Python readout/statistics tests passed. - All 73 gates, generated projections, strict documentation build and local aggregate passed (10 passed; one documented optional absence). - All nine PR checks succeeded at `7df0d42b12f35ffc22008cc10a40339afcfbb6a0`. - Fresh author-distinct review passed all six acceptance criteria over all 59 changed paths, with no findings or unchecked scope, after repairing the criterion-coverage finding. ## Evidence limits The real native coding repair demonstrates usability, not comparative skill uplift. The strict live-session machine replay remains NOT_PROVEN where execution/identity observations are unavailable; the source review PASS is retained separately. Existing cohort limits and the historical aggregate-enforcement gap remain unwaived. No new comparative cohort, scheduler, skill-corpus deletion, memory migration or global installation is included.
Role-Based Profiles
Purpose: Organize 12 domain skills into 3 discoverable profiles for different work contexts.
Version: 3.3.0 (Updated for skills-based approach)
How to Use Profiles
Profiles are documentation groupings, not executable configs. They organize AgentOps skills by domain so you can quickly find which skills are relevant to your work.
AgentOps 3.0 workflow recipes such as product-council,
engineering-team, and always-on-substrate were documented in
Activation Profiles (historical reference; retired path: ../activation-profiles.md). Activation profiles are
product recipes with explicit inputs, commands, and expected artifacts; this
page remains the role taxonomy.
To use a profile:
- Find your domain in the table below
- Open the profile YAML (e.g.,
docs/profiles/roles/software-dev.yaml) to see which skills it groups together - Use those skills directly via slash commands (
/research,/plan,/implement) or by reading the SKILL.md files listed in the profile
You do not "load" or "activate" a profile — you read it to discover which skills apply to your workflow, then invoke those skills individually.
Quick Start
| You're doing... | Profile | Key Skills |
|---|---|---|
| Building apps (APIs, frontends, features) | software-dev | languages, development, code-quality |
| Operations (incidents, monitoring, deploys) | platform-ops | operations, monitoring, security |
| Writing (docs, research, patterns) | content-creation | documentation, research, meta |
The 3 Profiles
1. Software Development (software-dev)
What you do:
- Build applications (backend APIs, frontends, full-stack)
- Write code in Python, Go, Rust, TypeScript, Java
- Review code and generate tests
- Deploy with CI/CD pipelines
Skills included:
| Skill | Knowledge Areas |
|---|---|
| languages | Python, Go, Rust, Java, TypeScript, Shell (6) |
| development | Backend, frontend, fullstack, mobile, iOS, deployment, AI, prompts (8) |
| code-quality | Review, improve, test generation (3) |
| validation | Assumptions, continuous, planning, tracer bullets (4) |
| data | Engineering, science, ML, MLOps (4) |
Example workflow:
1. /research "API design for user auth"
2. Load languages skill for Python patterns
3. Load development skill for backend architecture
4. Load code-quality skill before commit
2. Platform Operations (platform-ops)
What you do:
- Respond to production incidents
- Monitor system health and performance
- Debug errors and analyze logs
- Manage deployments and infrastructure
Skills included:
| Skill | Knowledge Areas |
|---|---|
| operations | Incident response, triage, postmortems, error detection (4) |
| monitoring | Alerts/runbooks, performance engineering (2) |
| security | Penetration testing, network engineering (2) |
| validation | Assumptions, tracer bullets (4) |
| meta | Context, execution, autonomous work (6) |
Example workflow:
1. Alert fires → Load operations skill
2. Check metrics → Load monitoring skill
3. After resolution → Load meta skill for postmortem
3. Content Creation (content-creation)
What you do:
- Write documentation (technical and non-technical)
- Extract patterns from completed work
- Conduct research and analysis
- Create tutorials and examples
Skills included:
| Skill | Knowledge Areas |
|---|---|
| documentation | Create, optimize, Diátaxis audit, API docs (4) |
| research | Code, docs, history, archive, structure, specs (6) |
| meta | Context, observer, memory, retro analysis (6) |
| specialized | Accessibility, support, UX (6) |
Example workflow:
1. /research "authentication patterns in our codebase"
2. Load research skill for exploration
3. Load documentation skill for writing
4. Load meta skill for retrospective
Skills by Profile
| Skill | software-dev | platform-ops | content-creation |
|---|---|---|---|
| languages | ✅ | ||
| development | ✅ | ||
| documentation | ✅ | ||
| code-quality | ✅ | ||
| research | ✅ | ||
| validation | ✅ | ✅ | |
| operations | ✅ | ||
| monitoring | ✅ | ||
| security | ✅ | ||
| data | ✅ | ||
| meta | ✅ | ✅ | |
| specialized | ✅ |
Loading Skills
Skills load into main context with full tool access:
# Read a skill directly
Read skills/languages/SKILL.md
# Or let triggers auto-load
"I need help with Python async patterns"
# → languages skill auto-activates
Profile Files
profiles/
├── README.md # This file
├── roles/
│ ├── software-dev.yaml # Software development profile
│ ├── platform-ops.yaml # Platform operations profile
│ └── content-creation.yaml # Content creation profile
└── examples/
├── software-dev-session.md # Example dev session
├── platform-ops-session.md # Example ops session
└── content-creation-session.md # Example writing session