Agent skill

aidd-core

Aggregates shared runtime navigation and cross-skill ownership links. Use when resolving shared runtime entrypoints or ownership boundaries.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/aidd-core

SKILL.md

Purpose

  • This skill is the thin runtime-oriented navigation layer.
  • Policy guidance (output contract, question format, read discipline, loop safety) lives in feature-dev-aidd:aidd-policy.
  • Domain owners are split into dedicated shared skills (aidd-docio, aidd-flow-state, aidd-observability, aidd-loop, aidd-rlm).
  • Agents may reference feature-dev-aidd:aidd-core as the topology entrypoint and navigation hub.

Canonical shared Python entrypoints

  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-rlm/runtime/rlm_slice.py
  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/diff_boundary_check.py
  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/plan_review_gate.py
  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/prd_review_gate.py
  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/rlm_targets.py
  • python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/skill_contract_validate.py

Command contracts

python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/skill_contract_validate.py

  • When to run: before release or CI checks when skill contract drift is suspected.
  • Inputs: repository root plus optional filters for prompt/skill validation scope.
  • Outputs: deterministic validation diagnostics for contract and prompt structure issues.
  • Failure mode: non-zero exit when required contract fields/sections are missing or inconsistent.
  • Next action: fix reported contract gaps in owners, then rerun the same validator.

python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-core/runtime/diff_boundary_check.py

  • When to run: during loop/review orchestration before applying actions that may cross boundaries.
  • Inputs: diff metadata (--ticket, --scope-key, optional work-item context).
  • Outputs: boundary verdict (ok|warn|blocked) and evidence references.
  • Failure mode: non-zero exit when diff inputs are missing or policy boundary checks fail.
  • Next action: tighten scope/actions or hand off blocker, then rerun the boundary check.

Ownership guard

  • aidd-core owns only aggregator-level shared runtime modules.
  • Policy/procedure text must be hosted in feature-dev-aidd:aidd-policy.

Additional resources

  • Policy contract: ../aidd-policy/SKILL.md (when: output/read/question rules are needed; why: apply shared policy without duplicating stage docs).
  • DocIO runtime owner: ../aidd-docio/SKILL.md (when: actions/map/markdown tooling is needed; why: use canonical DocIO entrypoints from a dedicated shared skill).
  • Flow/state runtime owner: ../aidd-flow-state/SKILL.md (when: stage state/progress/tasklist lifecycle tools are needed; why: use dedicated flow-state entrypoints).
  • Observability runtime owner: ../aidd-observability/SKILL.md (when: diagnostics/inventory/log/report commands are needed; why: keep observability surface under a dedicated shared owner).
  • Context pack template source: templates/context-pack.template.md (when: context pack format is unclear; why: confirm canonical sections generated by init/runtime).

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results