/ownership-agent

k9_ownership_preferences Schema Reference

Schema (12 columns, all STRING)

CSV Header

id,preference_type,tag_key,tag_value,owner,confidence,owner_type,handle,exclusion_type,exclusion_resource_type,prompt_text,priority

Each row gets a unique sequential id and fills columns relevant to its preference_type, leaving the rest empty.

Preference Types

Tag Mappings

A tag mapping says: "When a resource has tag X:Y, it belongs to this owner."

The agent checks cloud resource tags against your mappings. When a match is found, the specified owner is added as a candidate. Multiple mappings can match the same resource, producing multiple candidates ranked alongside other data sources.

Tag mappings complement existing data sources — they do not override a direct ownership tag (like dd-team) already on the resource.

Columns: id (required), preference_type=tag_mapping, tag_key (required), tag_value (optional, empty=wildcard), owner (required), confidence (required: high/medium/low), owner_type (required: team/user/service).

Owner type guidance:

Confidence guidance:

Matching behavior:

• Tag key and value matching is case-insensitive. Cost-Center matches cost-center.
• An empty tag_value matches any value for that tag key (wildcard).
• If multiple mappings match, all produce candidates. The agent ranks them by confidence.

Example rows:

1,tag_mapping,cost-center,CC-100,team-platform,high,team,,,,,
2,tag_mapping,managed-by,,team-infra,low,team,,,,,

Exclusions

An exclusion says: "Never assign this handle as a resource owner."

Bot accounts, CI runners, and shared service accounts often appear in cloud resource metadata. Exclusions remove these from ownership results.

Columns: id (required), preference_type=exclusion, handle (required), exclusion_type (optional), exclusion_resource_type (optional).

Matching behavior:

• The handle is matched case-insensitively.
• Optional filters use AND logic. All non-empty fields must match for the exclusion to apply.
• Leave exclusion_type and exclusion_resource_type empty to exclude from all results (most common).

Example rows:

1,exclusion,,,,,,deploy-bot,,,,
2,exclusion,,,,,,ci-runner,service,,,
3,exclusion,,,,,,k8s-node-controller,service,aws_ec2_instance,,

Custom Prompt Text

Custom prompt text provides free-form guidance to the AI inference engine. Use it to share organizational context: naming conventions, team structures, which data sources to prioritize.

Up to 3 entries, one per priority level (high, medium, low). Entries with the same priority are concatenated.

Columns: id (required), preference_type=prompt_text, prompt_text (required, up to 4096 bytes), priority (optional, default: low).

Tips for effective guidance:

• Be specific and actionable: "The cost-center tag is our most reliable ownership signal" > "Use tags"
• Use plain, declarative sentences — describe facts, not instructions to the AI
• Avoid special formatting: Markdown, HTML, XML tags are stripped during processing

Example rows:

1,prompt_text,,,,,,,,,Our organization assigns ownership by cost center.,high
2,prompt_text,,,,,,,,,Shared infrastructure accounts should never be resource owners.,medium

Validation Rules

All-or-nothing: If any row fails validation, the entire preference set is rejected for that sync cycle. Preferences are left empty until a valid set is uploaded.

Allowed Characters

Not allowed in any field: Angle brackets (< >), curly braces ({ }), pipe characters (|).

Size Limits

Duplicate Detection

The agent rejects the entire set if it contains conflicts:

• Tag mappings: Same tag_key+tag_value with different owner = conflict. Same key+value+owner with different confidence = conflict. Exact duplicates are allowed.
• Exclusions: Same handle+exclusion_type+exclusion_resource_type = duplicate. Case-insensitive.