autonomous-agent-engineering
Skill1 starsChanged 7 days agoMITView on GitHub
Core discipline, orchestration, and decision-making for autonomous AI coding agents. Enforces minimal diffs (Ponytail Engine), silent self-interrogation across 14 engineering domains (Socratic), multi-agent deliberation (Court & Arena), multi-model resilience (Model Router), durable memory vault, context budget management, and intent scoping.
In this repository62 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/skills/humanizer/AGENTS.mdAGENTS.md
- agentshieldSkill
- asd-ste100Skill
- awwwardsSkill
- backend-engineeringSkill
- boarduiSkill
- design-systemsSkill
- frontend-engineeringSkill
- interface-designSkill
- jurySkill
- karpathy-guidelinesSkill
- last30daysSkill
- lemo-opuscarSkill
- motionsites-web-builderSkill
- quality-engineeringSkill
- rev-engineerSkill
- scraplingSkill
- security-auditSkill
- skillforgeSkill
- stageshowSkill
- technical-documentationSkill
- agentshieldSkill
- anti-vcodeSkill
- arenaSkill
- asd-ste100Skill
- autonomous-agent-engineeringSkill
- awwwardsSkill
- backend-engineeringSkill
- boarduiSkill
- courtSkill
- designsSkill
- design-systemsSkill
- diagram-designSkill
- frontend-engineeringSkill
- humanizerSkill
- impeccableSkill
- interface-designSkill
- jurySkill
- karpathy-guidelinesSkill
- last30daysSkill
- lemo-opuscarSkill
- logo-designSkill
- model-routerSkill
- motionsites-web-builderSkill
- ponytailSkill
- quality-engineeringSkill
- rev-engineerSkill
- rigorSkill
- sanzo-wada-colorsSkill
- scraplingSkill
- secure-auditSkill
- security-auditSkill
- skillforgeSkill
- socraticSkill
- stageshowSkill
- stealSkill
- tasteSkill
- technical-documentationSkill
- threeuiSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 39
When proposals require rigorous critical pushback, simulate prosecution/defense debate or bracket tournaments.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Autonomous Agent Engineering
Consolidated capability governing agent execution discipline, minimal diff footprints, adversarial deliberation, multi-model resilience, durable context sharing, and context budget management.
When to Activate
- Scoping ambiguous or underspecified user requests into testable acceptance criteria.
- Enforcing minimal diffs, YAGNI, and standard library reuse on coding tasks.
- Interrogating system architecture silently across 14 engineering domains before planning.
- Stress-testing architectural proposals or technical claims via adversarial deliberation (Court / Arena).
- Managing multi-model failover cascades (429 handling) and task specialization.
- Preserving durable memory across agent sessions or managing context compaction.
When NOT to Activate
- Direct implementation of frontend React code (use
frontend-engineering). - Writing unit and E2E test suites (use
quality-engineering).
Core Execution Workflow
- Ponytail Minimal Diff & YAGNI Principle:
- Smallest complete diff: write the fewest changed lines, touch the fewest files, introduce zero speculative abstractions. Standard library first.
- Consult references/ponytail-yagni.md.
- Silent Pre-Implementation Interrogation (Socratic):
- Interrogate design silently across 14 engineering domains (requirements, API, security, infra, data, testing). Surface only decisions requiring human authority.
- Consult references/socratic-interrogation.md.
- Adversarial Deliberation (Court & Arena):
- When proposals require rigorous critical pushback, simulate prosecution/defense debate or bracket tournaments.
- Consult references/court-deliberation.md and references/arena-tournament.md.
- Multi-Model Resilience & Context Vault:
- Route algorithmic reasoning to deep thinking models; cascade on 429 rate limits.
- Maintain durable handoffs via memory vaults and compact context at natural task boundaries.
- Consult references/model-router.md, references/agent-unified-memory.md, and references/context-budget-compaction.md.
Modular Reference Directory
| Reference Module | Primary Focus | Source Provenance |
| ponytail-yagni.md | Lazy senior dev mode, minimal diffs, YAGNI, standard library first | hamzzaqureshi/gravidex (ponytail) |
| socratic-interrogation.md | Silent self-interrogation across 14 engineering domains | hamzzaqureshi/gravidex (socratic) |
| court-deliberation.md | Adversarial panel trial (prosecutor, defense, jury, verdict) | hamzzaqureshi/gravidex (court) |
| arena-tournament.md | Single-elimination subagent bracket tournament deliberation | hamzzaqureshi/gravidex (arena) |
| model-router.md | Task-specialized model routing and resilient 429 failover cascade | hamzzaqureshi/gravidex (model-router) |
| impeccable-lifecycle.md | 24-command design lifecycle system and UI architecture framework | hamzzaqureshi/gravidex (impeccable) |
| agent-unified-memory.md | Durable memory vault and handoffs across Claude, OpenCode, Antigravity | affaan-m/ECC (unified-memory) |
| agent-introspection-recovery.md | Structured agent self-debugging and failure recovery reports | affaan-m/ECC (agent-introspection-debugging) |
| context-budget-compaction.md | Context capacity checks, strategic compaction at logical phases | affaan-m/ECC (context-budget, strategic-compact) |
| intent-driven-scoping.md | Scoping ambiguous user requests into verifiable acceptance criteria | affaan-m/ECC (intent-driven-development) |
| plan-canvas-annotations.md | Interactive local browser canvas for plan review and annotations | affaan-m/ECC (plan-canvas) |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.