← all publishers

kodexa-ai

@kodexa-ai source repo

Report this profile Claim this profile

• 19 skills
• 0 followers
• 7 hours ago last updated

1. ▌
   Label · kodexa-ai

   Use when creating, syncing or debugging Kodexa labels — the org-scoped tags Studio calls Document Tags — covering the missing slug that makes `kdx apply` refuse a label file, the name-matching rules that decide whether a push updates or silently no-ops, and the four runtime paths that create label rows behind your back
   0
   installs
2. ▌
   Store · kodexa-ai bundle

   Use when authoring or editing a Kodexa document store or data store — the org-scoped containers documents and extracted data live in. Covers the flattened wire shape and the inner keys that vanish if you write them flat, storeType and storePurpose values, the legacy type: store remapping, and the binding a store needs before anything can see it.
   0
   installs
3. ▌
   Intake · kodexa-ai bundle

   Use when documents need a way into Kodexa — authoring intakes/*.yaml, pointing an intake at a document store, wiring it to an activity plan, writing the Goja upload script, issuing upload tokens, choosing a sourceType, or diagnosing an INTAKE_* upload rejection.
   0
   installs
4. ▌
   Module · kodexa-ai bundle

   Use when authoring or debugging a Kodexa module.yml / model.yml — Python, Go-WASM or inline-JavaScript model modules and agent skill packs. Covers moduleType, moduleRuntimeRef, moduleSidecars, inferenceOptions, actions, contents/ignoredContents packaging, the parameters the runtime injects, and applying with kdx.
   0
   installs
5. ▌
   Kdx CLI · kodexa-ai bundle

   Use when running the Kodexa CLI (kdx) — logging in and managing profiles, listing/describing/deleting resources, validating and applying resource YAML, invoking declared API operations with run, pulling/pushing/deploying metadata with a sync manifest, inspecting local KDDB documents, and the project/store/task/secret/intake/document-family/knowledge command groups.
   0
   installs
6. ▌
   Trigger · kodexa-ai

   Use when creating, editing, or debugging Kodexa triggers — project-scoped YAML rules that start an activity plan when a platform event fires. Covers which event kinds actually dispatch, the exact payload each one carries, JSONata eventFilter and inputMapping, and the project binding a trigger needs before it can fire.
   0
   installs
7. ▌
   Assistant · kodexa-ai

   Use when creating, editing or reviewing a Kodexa assistant — the project-scoped record that holds options.pipeline steps, taxonomy refs and agent config, and that supplies the identity a project's executions run as. Also read this before wiring anything event-driven or scheduled, because that is authored as an activity-plan plus a trigger, not as an assistant.
   0
   installs
8. ▌
   Data Form · kodexa-ai bundle

   Use when creating or editing a Kodexa data form — the schema-driven review panel that shows and corrects extracted document data. Covers the V2 `nodes` tree and its `v2:*` components, legacy V1 `cards`, `tagPath` binding, `entrypoints`, form scripts and the bridge, keyboard shortcuts and declarative tab order.
   0
   installs
9. ▌
   Task Status · kodexa-ai bundle

   Use when defining or changing the workflow states Kodexa tasks move through — authoring task-statuses/*.yaml, picking statusType (OPEN / IN_PROGRESS / PENDING / BLOCKED / DONE), configuring locked and lockDocumentFamily, retiring a state, or binding org-level statuses to a project so reviewers can reach them.
   0
   installs
10. ▌
    Channel Type · kodexa-ai bundle

    Use when configuring the Kodexa Studio chat agent — choosing which MCP servers, built-in skills, module skill packs and system-prompt fragment a chat loads, editing a channel-type row, or debugging a chat that has no platform tools at all. Covers the metadata payload, the strict opt-in mount rules, and the bind-once channel binding.
    0
    installs
11. ▌
    Activity Plan · kodexa-ai bundle

    Use when writing or editing Kodexa ActivityPlan YAML — the org-scoped graph of steps (EXECUTION, CREATE_TASK, SCRIPT, LLM, BRIDGE_CALL, AGENT) that runs as a project Activity. Covers the flat step envelope keyed by `type`, dependsOn and action edges, per-document fan-out and routing (perDocument, ANY_BRANCH, await `?` deps), setDocumentStatus, inputOptions vs inputsSchema, per-field reference formats, and the several template languages that coexist in one plan.
    0
    installs
12. ▌
    Task Template · kodexa-ai bundle

    Use when creating or editing a Kodexa task template — the org-scoped YAML behind a human review/approval task: action buttons and their status transitions, attached data forms, document upload groups, AI task naming, chat prompt, agent shortcuts, panel and button visibility, team/priority defaults, and task locking. Also use when an activity-plan CREATE_TASK step references a task template by slug.
    0
    installs
13. ▌
    Service Bridge · kodexa-ai bundle

    Use when authoring, reviewing or debugging a Kodexa service bridge — the org-scoped YAML that proxies an external HTTP API: baseUrl, named endpoints, defaultHeaders carrying ${secrets.NAME}, OAuth2 client-credentials auth, per-endpoint response caching, requestSchema guards, initScript/preSendScript/postReplyScript hooks, and the agentCallable flag that lets activity-plan BRIDGE_CALL steps and agents invoke it.
    0
    installs
14. ▌
    Data Definition · kodexa-ai bundle

    Use when creating, editing, or debugging a Kodexa data-definition (taxonomy) — the YAML that defines taxons, data types, extraction prompts (semanticDefinition), typeFeatures, repeating groups and line items, KEXL formulas and validation rules, conditional formatting, selection options and event subscriptions. Covers data-definition:// and taxonomy:// resources applied with kdx.
    0
    installs
15. ▌
    Prompt Template · kodexa-ai bundle

    Use when creating or editing a Kodexa prompt resource — flat YAML with type: prompt-template in kdxa_prompts, carrying promptTemplate + templateType (FSTRING or MUSTACHE) and/or a plain prompt body, referenced by activity-plan LLM steps via promptTemplateRef, by the organization LLM endpoint via promptRef, by Python via Prompt.render(), and by the Studio chat prompt picker. Covers which key each consumer reads, brace syntax, and the fields that do not exist.
    0
    installs
16. ▌
    Knowledge System · kodexa-ai bundle

    Use when authoring or debugging Kodexa knowledge YAML — knowledge sets, knowledge feature types, knowledge item types, feature instances and knowledge items — including featureExpression trees, content-addressable feature slugs, option schemas, attachment:// references, set status and promotion, and diagnosing a set that never matches a document.
    0
    installs
17. ▌
    Project Resource · kodexa-ai

    Use when making an org-scoped Kodexa resource usable inside a project — binding activity-plans, task-templates, task-statuses, data-definitions, data-forms, stores, prompts, service-bridges or intakes, writing `linked:` manifest blocks or project-template ref arrays, or debugging "not bound to project", empty pickers and 409-on-delete
    0
    installs
18. ▌
    Project Template · kodexa-ai bundle

    Use when authoring or editing a Kodexa project-template YAML — the org-scoped blueprint that provisions a new project's stores, assistants, taxonomies, data forms, document and task status workflows, task templates, knowledge sets, activity-plan bindings, triggers, project options and parent/child lineage config
    0
    installs
19. ▌
    Metadata Envelope · kodexa-ai bundle

    Use when authoring or debugging any Kodexa resource YAML — the shared slug / name / orgSlug / type envelope twelve org-scoped resource types embed, the flatten rule that makes authored YAML flat, what a slug is actually checked against, the changeSequence optimistic lock, the computed ref and uri, and where each file lives in a metadata repo.
    0
    installs