aoa-techniques

Public library of reusable techniques for coding agents and humans.

aoa-techniques is the public practice canon of AoA. It is not a snippet dump and not an “awesome list”. A technique here is a minimal reproducible unit of engineering practice: a workflow, validation pattern, safety protocol, documentation layout, evaluation loop, or transfer method that can travel cleanly across projects.

Start here

Use the shortest route by need:

• repo-owned entrypoint: docs/START_HERE.md
• one full bundle end to end: techniques/agent-workflows/plan-diff-apply-verify-report/TECHNIQUE.md
• current technique map: TECHNIQUE_INDEX.md
• verify current repo state: python scripts/validate_repo.py and python -m unittest discover -s tests
• release-prep parity path: docs/RELEASING.md, python scripts/release_check.py, and git status -sb
• deeper docs map: docs/README.md
• authoring template: templates/TECHNIQUE.template.md
• contribution path: CONTRIBUTING.md

Quick routes

• repo layer position and neighboring repos: docs/ECOSYSTEM_CONTEXT.md
• current technique map, docs map, and generated catalog: TECHNIQUE_INDEX.md, docs/README.md, and generated/technique_catalog.min.json
• feat-reader and runtime-card surfaces: docs/TECHNIQUE_FEAT_MODEL.md, generated/technique_feat_cards.min.example.json, docs/TECHNIQUE_CAPSULES.md, and generated/technique_capsules.min.json
• status, review, and promotion posture: docs/CANONICAL_RUBRIC.md, docs/CANONICAL_REVIEW_GUIDE.md, docs/PROMOTION_READINESS_MATRIX.md, and docs/RELEASING.md
• one end-to-end example path: WALKTHROUGH.md
• bounded execution and proof neighbors: aoa-skills and aoa-evals
• navigation and ecosystem map: aoa-routing and Agents-of-Abyss

Deeper routes

• donor intake, refinement, and promotion: docs/DONOR_REFINERY_RUBRIC.md, docs/EXTERNAL_IMPORT_RUNBOOK.md, docs/CROSS_LAYER_TECHNIQUE_CANDIDATES.md, docs/PROMOTION_READINESS_MATRIX.md, docs/PROMOTION_WAVE_A_RUNBOOK.md, docs/EXTERNAL_EVIDENCE_SPRINT_RUNBOOK.md, and docs/EXTERNAL_EVIDENCE_LEDGER.md
• selection and chooser surfaces: docs/TECHNIQUE_SELECTION_GUIDE.md, docs/TECHNIQUE_SELECTION.md, and docs/SELECTION_PATTERNS.md
• runtime cards and capsule surfaces: docs/TECHNIQUE_CAPSULES.md, generated/technique_capsules.json, generated/technique_capsules.min.json, and docs/TECHNIQUE_CAPSULE_GUIDE.md
• repo-doc routing and authoritative doc/status manifests: docs/REPO_DOC_SURFACES.md, generated/repo_doc_surface_manifest.json, and docs/REPO_DOC_SURFACE_LIFT_GUIDE.md
• source-lift, KAG, and section families: docs/KAG_EXPORT.md, generated/kag_export.json, generated/kag_export.min.json, docs/KAG_SOURCE_LIFT_GUIDE.md, docs/TECHNIQUE_SECTIONS.md, generated/technique_sections.full.json, docs/TECHNIQUE_CHECKLISTS.md, docs/TECHNIQUE_EXAMPLES.md, docs/EVIDENCE_NOTE_SURFACES.md, generated/technique_section_manifest.json, generated/technique_checklist_manifest.json, generated/technique_example_manifest.json, generated/technique_evidence_note_manifest.json, docs/TECHNIQUE_SECTION_LIFT_GUIDE.md, docs/TECHNIQUE_CHECKLIST_LIFT_GUIDE.md, docs/TECHNIQUE_EXAMPLE_LIFT_GUIDE.md, and docs/EVIDENCE_NOTE_PROVENANCE_GUIDE.md
• review and shadow surfaces: docs/SHADOW_PATTERNS.md, docs/PUBLISHED_SUMMARY_SHADOW_REVIEW.md, docs/EVALUATION_CHAIN_SHADOW_REVIEW.md, generated/shadow_review_manifest.json, generated/semantic_review_manifest.json, and docs/KAG_SOURCE_LIFT_SEMANTIC_REVIEW.md
• owner-local live receipt publication for closeout/stats integration: scripts/publish_live_receipts.py and .aoa/live_receipts/technique-receipts.jsonl
• current verify and release-prep path: python scripts/validate_repo.py, python -m unittest discover -s tests, docs/RELEASING.md, python scripts/release_check.py, and git status -sb

What belongs here

Good candidates:

• agent workflows
• validation patterns
• documentation structures
• evaluation and monitoring loops
• safety and sanitization patterns
• infra operation techniques
• cross-repo promotion and reuse patterns

Bad candidates:

• random snippets
• private project hacks without adaptation notes
• secret-bearing configs
• raw logs
• project-only dumps
• undocumented scripts
• objects that belong as skills, evals, routing logic, role contracts, or playbooks

Core principles

• public-safe reusable practice over project-local residue
• bounded, reviewable contracts over vague lore
• source-linked promotion over raw copying
• linked docs and generated surfaces over oversized root inventories

Maturity model

This repository primarily exposes public techniques in promoted, canonical, and deprecated states. Earlier incubation may happen elsewhere, but the root here should point to curated public canon rather than replaying the full intake history.

Repository structure

• techniques/ for published technique bundles
• templates/ for technique authoring and promotion scaffolds
• generated/ for derived catalogs, capsules, source-lift, and review surfaces
• docs/ for doctrine, intake, release, selection, and shadow surfaces
• scripts/ and tests/ for validation and generation helpers

Intended users

• coding agents
• solo builders
• infra and product engineers
• teams building reusable operational knowledge

What a good technique includes

A strong technique should include:

• clear intent and usage boundaries
• explicit inputs, outputs, and risks
• validation method
• adaptation notes when portability needs them
• enough structure to be promoted, reviewed, and reused

Contribution model

aoa-techniques owns practice meaning, while neighboring repos own execution, proof, routing, role, and scenario composition. If a reusable contract can be extracted cleanly from a neighboring layer, it belongs here once it becomes public-safe, bounded, and portable.

The current runtime path for public technique use remains:

pick -> inspect -> expand -> object use

License

Apache-2.0