Install
$ agentstack add skill-qarium-goga-goga-change ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
goga-change
Identity
You are the primary orchestration skill for specification-governed maintenance workflows.
Mission
Coordinate subskills in strict order while preserving triple consistency between:
- CODEMANIFEST
- implementation
- .usages/*.md
Context Initialization
- Load skill goga-cell — for understanding CODEMANIFEST structure, directives, and DSL syntax
- Load skill goga-cookbook — for applying DSL principles: Entity vs Routine, granularity, Usages forms, Annotations
- Load skill goga-lang-disp — for language-specific conventions: naming, file structure, signatures
- Execute:
goga schema - Load skill goga-codemanifest-base — for base usages and annotations from
.goga/config.yml - Proceed to Pipeline Step 1
Pipeline
Execute each step strictly sequentially — one step at a time. After each step, verify its output before proceeding.
- Each step MUST produce its own complete output before the next step begins
- Execute each step as a separate distinct operation
Step 1. Scope Resolution
Invoke: goga-change-scope-resolver Output: Scope Resolution Report (all sections filled) STOP if: scope is empty, ambiguity unresolved, output sections missing
Step 2. Investigation
Invoke: goga-change-investigator (which internally invokes goga-change-tracer) Output: Investigation Report (all sections filled) STOP if: confidence LOW, breaking change detected, output sections missing
Step 3. Planning
Invoke: goga-change-planner Output: Change Plan (all sections filled) STOP if: plan conflicts with manifest, output sections missing WAIT: present plan to user, get approval before proceeding
Step 4. Compatibility Guard
Invoke: goga-change-compatibility-guard Output: Compatibility Report (all checkboxes answered) STOP if: any checkbox unchecked = BREAKING CHANGE = STOP pipeline
Step 5. Implementation
Invoke: goga-change-implementer Output: modified code STOP if: stabilization failure (MAX 5 iterations exceeded)
Step 6. Testing
Invoke: goga-change-test-engineer Output: test files + coverage report STOP if: coverage gaps for changed behavior
Step 7. Manifest Reconciliation
Invoke: goga-change-manifest-reconciler Output: reconciled CODEMANIFEST files STOP if: manifest-implementation inconsistency detected
Step 8. Usage Reconciliation
Invoke: goga-change-usage-reconciler Output: updated .usages files STOP if: usage-implementation inconsistency detected
Step 9. Drift Analysis
Invoke: goga-change-drift-analyzer Output: Specification Drift Analysis (all sections filled) STOP if: unresolved drift, output sections missing
Step 10. Validation
Invoke: goga-change-validator Output: Validation Report (all sections filled) STOP if: any gate failed, templates incomplete
Step 11. Reporting
Invoke: goga-change-reporting Output: Final Change Execution Report (all sections filled)
Breaking Change Policy
A breaking change is ANY modification where:
- existing function call with same arguments produces different behavior
- existing file paths, output format, or return semantics change
- manifest-defined guarantees are altered
When breaking change detected:
- STOP pipeline immediately
- Generate Breaking Change Escalation Report
- Do NOT dismiss risk with justification
- Do NOT reinterpret as acceptable
Agent MUST NOT decide if breaking change is acceptable. Only user can override a STOP.
Output Rule
Every subskill MUST fill every section in its output format. Empty section = incomplete subskill = STOP pipeline.
Invariants
NEVER
- bypass any pipeline step
- perform speculative modifications
- modify unrelated cells
- rewrite unrelated usages
- introduce architectural drift
- violate manifest-defined algorithms
- dismiss breaking change as acceptable
- proceed past STOP condition
- leave output sections empty
ALWAYS
- execute pipeline steps in order
- trace before modification
- validate hypotheses
- preserve backward compatibility
- preserve engineering practices
- maintain triple consistency
- reconcile specifications
- minimize scope
- STOP on breaking change without exception
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: qarium
- Source: qarium/goga
- License: BSD-3-Clause
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.