Files
PaperClipAI/packages/shared/src/validators/responsibility.ts
T
575ef12575 feat(responsibilities): file company skills under responsibilities (#22275)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - A responsibility shows the skills of the agents doing its work.
People cannot say which skills a responsibility needs.
> - The maintainer chose to file skills under responsibilities like
projects, tasks and routines.
> - Agents also need to set a skill's catalog details without rewriting
its SKILL.md.
> - This pull request lets people and agents file, move and remove
company skills under a responsibility, and adds `update_skill_details`.
> - The benefit is that a responsibility lists the skills it needs,
beside the skills its agents already hold.

## Linked Issues or Issue Description

Refs #22253, #22261 and #22274; this pull request is stacked on them.

**Subsystem affected**

Database (`responsibility_links`), shared types and validators, server
responsibility service and routes, Dashboard versions, native runner
tools, and the responsibility page UI.

**Problem or motivation**

Skills under a responsibility are only computed from the agents doing
its work. A person cannot record that "Hiring needs interviewing", and
an agent cannot file a skill when it maps work for a person.

**Proposed solution**

Add a nullable `skill_id` to `responsibility_links` and allow it as the
one target of a link. Show filed and computed skills together. Let the
file, unfile, lookup and link routes and tools take skills. Keep filed
skills in Dashboard versions so undo covers them. Add a small live tool
for a skill's tagline and categories.

**Alternatives considered**

Keeping skills computed only, and giving the skill to the agent that
runs the responsibility, was the other option. It was not chosen.
Extending `update_skill` would change its scenario contract, whose mock
needs `markdown`, so the details tool is separate.

## What Changed

- Review fixes: the default responsibility view's Skills "+" adds an
existing skill, removes a filed one, or opens an agent's skills page. A
newest Dashboard version without a skill list is upgraded in place, so
the first skill filed after it is undone. Undo files again, or removes,
only skills the restoring person can see. A read scope without a skill
condition sees no filed skill. `update_skill_details` with only
`skillId` is a no-op, as its schema allows.

- Migration `0499`: add `responsibility_links.skill_id` (FK to
`company_skills`, on delete cascade). Widen the single-target check to
four targets. Add a unique `(responsibility_id, skill_id)` index and a
`skill_id` index. The migration is idempotent and online-compatible (NOT
VALID then VALIDATE, and the indexes are built CONCURRENTLY). It is
listed in `MIGRATIONS_WITH_TOP_LEVEL_STATEMENTS`.
- Shared: `ResponsibilityLink.skillId`;
`ResponsibilitySkillSummary.linkId` and `computed`; `skillId` in the
link and lookup schemas; `skill` in `RESPONSIBILITY_FILE_ITEM_KINDS`;
optional `skillIds` in Dashboard snapshot links.
- Server: filing checks the reader's skill visibility
(`skillAccessService`). The detail merges filed and computed skills into
one card each. Lookup accepts `skillId`. Dashboard capture, restore and
summary include filed skills. A version without `skillIds` leaves filed
skills as they are when it is restored.
- Tools: `file_under_responsibility`, `remove_from_responsibility` and
`find_responsibilities_for` accept `kind: "skill"`. `get_responsibility`
returns `linkId` and `computed` for each skill. Add
`update_skill_details` (tagline and categories).
- UI: the responsibility page has "Add existing skill". A filed skill
shows "Filed here" and a remove button. Computed skills keep their
holders.

## Verification

- Server: the `responsibilit*` tests (new: file, find, move and unfile a
skill; one card for a skill that is both filed and computed; undo of a
filed skill from the Dashboard history), `skill-tools.test.ts` (new
`update_skill_details` cases), `responsibility-tools`, `routine-tools`,
`routines-routes`, `paperclip-runner-tool-authority`, `runner-api` and
`openapi-routes`: 24 files, 2723 tests passed.
- UI: responsibility page and component tests (16 files, 158 tests,
including a new filed-skill test) pass. `pnpm --filter @paperclipai/ui
typecheck` passed.
- `@paperclipai/db` `check:migrations` passes and reports the override.
The snapshot drift test passes. `tsc` passes for `db`, `shared` and the
server layers.
- Runner: catalog and semantic-tool tests (317) pass. All generator
checks pass.

## Risks

- The migration changes a check constraint on `responsibility_links`.
The replacement accepts every row that the old check accepted. It is
swapped as NOT VALID and then validated. The previous release never sets
`skill_id`.
- Deleting a company skill removes its links (cascade).
- Filed skills follow the reader's skill visibility. A skill the reader
may not see is not listed.

## Model Used

Claude Opus 5.5 (`claude-opus-5-5`) in Claude Code, with tool use and
code execution.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-11-19 09:30:49 -06:00

280 lines
12 KiB
TypeScript

import { z } from "zod";
import {
RESPONSIBILITY_AREA_SOURCES,
RESPONSIBILITY_AREA_STATUSES,
RESPONSIBILITY_DELEGATION_MODES,
RESPONSIBILITY_HEALTH_OVERRIDES,
RESPONSIBILITY_METRIC_DIRECTIONS,
RESPONSIBILITY_SOURCES,
RESPONSIBILITY_STATUSES,
} from "../constants.js";
import {
RESPONSIBILITY_METRIC_KINDS,
resolveResponsibilityMetricKind,
responsibilityBarIssues,
responsibilityMetricNameSchema,
responsibilityMetricReadingSchema,
responsibilityMetricUnitSchema,
responsibilityMetricValueSchema,
} from "../responsibility-bars.js";
const title = z.string().trim().min(1).max(200);
const longText = z.string().trim().max(10_000);
const optionalText = longText.optional().nullable();
const requiredText = longText.min(1);
const metricName = responsibilityMetricNameSchema;
const metricUnit = responsibilityMetricUnitSchema;
const metricValue = responsibilityMetricValueSchema;
const sortOrder = z.number().int().min(-1_000_000).max(1_000_000);
/** The inputs that set a metric on create: sending any one of them means the whole metric must be there. */
const CREATE_METRIC_KEYS = [
"metricKind",
"metricName",
"metricUnit",
"metricDirection",
"metricTarget",
"metricLow",
"metricHigh",
"metricCurrent",
] as const;
const createResponsibilityFields = z.object({
title,
description: optionalText,
/**
* The bar, as a sentence. A person may leave it out (with the metric) and
* set the bar later; it is required with a metric, and always from an agent.
*/
standard: optionalText,
/**
* The bar template. Omitted: `percentage` when metricUnit is "%", otherwise
* `volume`. Each template requires and allows its own fields (see
* `RESPONSIBILITY_BAR_TEMPLATES`).
*/
metricKind: z.enum(RESPONSIBILITY_METRIC_KINDS).optional(),
/** What is measured. A person may leave the whole metric out; required from an agent. */
metricName: metricName.optional(),
/** Volume and percentage (required for them). */
metricTarget: metricValue.optional().nullable(),
/** Range (required for it): the lowest value that is still inside. */
metricLow: metricValue.optional().nullable(),
/** Range (required for it): the highest value that is still inside. */
metricHigh: metricValue.optional().nullable(),
/** A number; true or false for `yes_no`. */
metricCurrent: responsibilityMetricReadingSchema.optional().nullable(),
metricUnit: metricUnit.optional().nullable(),
/** Volume and percentage only. Default `at_least`. */
metricDirection: z.enum(RESPONSIBILITY_METRIC_DIRECTIONS).optional(),
/** An area in the same company. */
areaId: z.string().uuid().optional().nullable(),
/**
* An area by title instead of `areaId`: the live area with that title (any
* case), or a new one made in the same transaction. Not with `areaId`.
*/
areaTitle: title.optional(),
/** An agent in the same company that runs the work. */
runByAgentId: z.string().uuid().optional().nullable(),
evidence: optionalText,
/** Board: defaults to the acting user. Agent: defaults to the person its run works for. */
ownerUserId: z.string().trim().min(1).optional(),
/** Board only; agents always create `proposed`. Board default: `active`. */
status: z.enum(RESPONSIBILITY_STATUSES).optional(),
/** Board only; agents always create `onboarding`. Board default: `manual`. */
source: z.enum(RESPONSIBILITY_SOURCES).optional(),
sortOrder: sortOrder.optional(),
});
type CreateResponsibilityFields = z.infer<typeof createResponsibilityFields>;
/** A metric that is sent must be whole: the sentence, what is measured, and the template's inputs. */
function refineCreateBar(value: CreateResponsibilityFields, ctx: z.RefinementCtx) {
const measured = CREATE_METRIC_KEYS.some((key) => value[key] !== undefined && value[key] !== null);
if (!measured) return;
if (!value.standard?.trim()) {
ctx.addIssue({ code: "custom", path: ["standard"], message: "A responsibility with a metric needs its bar as a sentence (standard)" });
}
if (!value.metricName?.trim()) {
ctx.addIssue({ code: "custom", path: ["metricName"], message: "A responsibility with a metric needs what is measured (metricName)" });
}
const metricKind = value.metricKind ?? resolveResponsibilityMetricKind(null, value.metricUnit);
for (const issue of responsibilityBarIssues({ ...value, metricKind }, { complete: true })) {
ctx.addIssue({ code: "custom", path: [issue.field], message: issue.message });
}
}
/**
* Create a responsibility. A person may send only a title: it is saved with
* no bar and reads as not measured until its bar is set. Any metric input
* makes the whole bar required (the sentence, what is measured and the
* template's inputs), so a half-filled metric is refused. Agents send
* `proposeResponsibilitySchema` instead, which always requires the bar.
*/
export const createResponsibilitySchema = createResponsibilityFields.strict().superRefine(refineCreateBar);
export type CreateResponsibility = z.infer<typeof createResponsibilitySchema>;
/** What an agent sends to propose a responsibility: the create body, with the whole bar required. */
export const proposeResponsibilitySchema = createResponsibilityFields
.extend({ standard: requiredText, metricName })
.strict()
.superRefine(refineCreateBar);
export type ProposeResponsibility = z.infer<typeof proposeResponsibilitySchema>;
export const updateResponsibilitySchema = z.object({
title: title.optional(),
description: optionalText,
/** `null` clears it; refused on an active responsibility. */
standard: optionalText,
/** `null` clears it; refused on an active responsibility. */
metricName: metricName.optional().nullable(),
/**
* Change the bar template. Inputs the new template does not use are cleared,
* and so is the reading unless metricCurrent is sent too. The merged bar
* must follow the template's rules (422 otherwise).
*/
metricKind: z.enum(RESPONSIBILITY_METRIC_KINDS).optional(),
/** `null` clears it; refused on an active responsibility. */
metricTarget: metricValue.optional().nullable(),
/** Range: `null` clears it; refused on an active responsibility. */
metricLow: metricValue.optional().nullable(),
/** Range: `null` clears it; refused on an active responsibility. */
metricHigh: metricValue.optional().nullable(),
/**
* The latest reading; sets `metricUpdatedAt`. A number, or true/false for a
* `yes_no` bar. On an active responsibility an agent may send this field
* alone when it runs the responsibility or is assigned a task or routine
* under it.
*/
metricCurrent: responsibilityMetricReadingSchema.optional().nullable(),
metricUnit: metricUnit.optional().nullable(),
/** Volume and percentage only. */
metricDirection: z.enum(RESPONSIBILITY_METRIC_DIRECTIONS).optional().nullable(),
areaId: z.string().uuid().optional().nullable(),
/** Move it to the live area with this title (any case), made in the same transaction when there is none. Not with `areaId`. */
areaTitle: title.optional(),
runByAgentId: z.string().uuid().optional().nullable(),
evidence: optionalText,
sortOrder: sortOrder.optional(),
/** `null` clears the override. */
healthOverride: z.enum(RESPONSIBILITY_HEALTH_OVERRIDES).nullable().optional(),
healthNote: optionalText,
status: z.enum(RESPONSIBILITY_STATUSES).optional(),
}).strict();
export type UpdateResponsibility = z.infer<typeof updateResponsibilitySchema>;
export const linkResponsibilitySchema = z.object({
projectId: z.string().uuid().optional(),
issueId: z.string().uuid().optional(),
routineId: z.string().uuid().optional(),
/** A company skill in the same company. */
skillId: z.string().uuid().optional(),
}).strict().refine(
(value) => [value.projectId, value.issueId, value.routineId, value.skillId].filter((id) => id !== undefined).length === 1,
{ message: "Provide exactly one of projectId, issueId, routineId or skillId" },
);
export type LinkResponsibility = z.infer<typeof linkResponsibilitySchema>;
export const delegateResponsibilitySchema = z.object({
title,
instructions: longText.optional().nullable(),
assigneeAgentId: z.string().uuid(),
mode: z.enum(RESPONSIBILITY_DELEGATION_MODES),
projectId: z.string().uuid().optional().nullable(),
}).strict();
export type DelegateResponsibility = z.infer<typeof delegateResponsibilitySchema>;
const booleanQuery = z
.union([z.boolean(), z.enum(["true", "false", "1", "0"])])
.transform((value) => value === true || value === "true" || value === "1");
export const listResponsibilitiesQuerySchema = z.object({
/** Board: defaults to the acting user. Agent: defaults to the person its run works for. */
ownerUserId: z.string().trim().min(1).optional(),
includeLetGo: booleanQuery.optional().default(false),
});
export type ListResponsibilitiesQuery = z.infer<typeof listResponsibilitiesQuerySchema>;
export const createResponsibilityAreaSchema = z.object({
title,
description: optionalText,
sortOrder: sortOrder.optional(),
/** Board only; agents always create `proposed`. Board default: `active`. */
status: z.enum(["proposed", "active"]).optional(),
/** Board only; agents always create `onboarding`. Board default: `manual`. */
source: z.enum(RESPONSIBILITY_AREA_SOURCES).optional(),
}).strict();
export type CreateResponsibilityArea = z.infer<typeof createResponsibilityAreaSchema>;
export const updateResponsibilityAreaSchema = z.object({
title: title.optional(),
description: optionalText,
sortOrder: sortOrder.optional(),
status: z.enum(RESPONSIBILITY_AREA_STATUSES).optional(),
}).strict();
export type UpdateResponsibilityArea = z.infer<typeof updateResponsibilityAreaSchema>;
export const listResponsibilityAreasQuerySchema = z.object({
includeArchived: booleanQuery.optional().default(false),
});
export type ListResponsibilityAreasQuery = z.infer<typeof listResponsibilityAreasQuerySchema>;
export const responsibilityLinkLookupQuerySchema = z.object({
issueId: z.string().uuid().optional(),
projectId: z.string().uuid().optional(),
routineId: z.string().uuid().optional(),
skillId: z.string().uuid().optional(),
}).refine(
(value) => [value.projectId, value.issueId, value.routineId, value.skillId].filter((id) => id !== undefined).length === 1,
{ message: "Provide exactly one of issueId, projectId, routineId or skillId" },
);
export type ResponsibilityLinkLookupQuery = z.infer<typeof responsibilityLinkLookupQuerySchema>;
/** What can be filed under a responsibility by kind and id. */
export const RESPONSIBILITY_FILE_ITEM_KINDS = ["project", "task", "routine", "skill"] as const;
export type ResponsibilityFileItemKind = (typeof RESPONSIBILITY_FILE_ITEM_KINDS)[number];
const fileItem = z.object({
kind: z.enum(RESPONSIBILITY_FILE_ITEM_KINDS),
id: z.string().uuid(),
}).strict();
/**
* File items under a responsibility in one step. `moveFrom` makes it a move:
* the items are also unfiled from that responsibility, or with `all` from
* every other proposed or active responsibility of the same owner.
*/
export const fileResponsibilityItemsSchema = z.object({
items: z.array(fileItem).min(1).max(50),
moveFrom: z.union([z.string().uuid(), z.literal("all")]).optional(),
}).strict();
export type FileResponsibilityItems = z.infer<typeof fileResponsibilityItemsSchema>;
/** Unfile items from a responsibility by kind and id; no link ids needed. */
export const unfileResponsibilityItemsSchema = z.object({
items: z.array(fileItem).min(1).max(50),
}).strict();
export type UnfileResponsibilityItems = z.infer<typeof unfileResponsibilityItemsSchema>;
/** An agent deletes only with the id of the confirmation card the person accepted. */
export const deleteResponsibilityQuerySchema = z.object({
confirmationId: z.string().uuid().optional(),
});
export const deleteResponsibilityAreaQuerySchema = z.object({
confirmationId: z.string().uuid().optional(),
/** An area for the responsibilities under it; without one they are left with no area. */
moveResponsibilitiesTo: z.string().uuid().optional(),
});