github-tools

Add an issue or pull request to a GitHub project

github_project_item_add

Idempotent

Add an issue or pull request to a GitHub project board and set field values in the same call; adding an existing item returns it without duplicate creation.

Instructions

Put an issue or pull request on a project board the viewer may update, and set field values in the same call. Adding is idempotent: one that is already on the board (archived or not) is returned as it is (created: false) and only the given fields are set. A new item with no Status given starts as Waiting when it is an issue with an open blocking issue and the board has a Waiting option, else as Todo when the board has that option; a board without a Status single select gets no default. Once the item is on the board, a field that fails to set does not fail the call: it is listed in errors as fields., and the item is still returned. If the blocking issues cannot be read the item starts as Todo and errors.blocked_by says why. Field values are set by field name and option or iteration name, matched without regard to case: text as a string, number as a number or numeric string, date as YYYY-MM-DD, single select as an option name, iteration as an iteration title or @current (the iteration containing today, UTC). Only text, number, date, single select and iteration fields can be set; a column that mirrors an issue field is set with the issue tools' issue_fields, not here. GitHub's own (Title, Assignees, Labels, Milestone, Repository, Linked pull requests, …) are set on the issue, and github_issue_edit sets assignees, labels and milestone. Every field and value is checked before anything is written. A board is named by project: its number, owner/number or its URL, used as it is; any other project is a title, and no project is Roadmap: the open board of that whole title is looked up, never guessed, in owner's boards if owner is given, else in the boards linked to repo, then its organization's, then the viewer's. Needs a token with the project scope; if GitHub reports missing scopes, run gh auth refresh -s project.

Input Schema

Name	Required	Description
`item`	Yes	The issue or pull request to put on the board: 12, #12, owner/repo#12 or its URL; a bare number means repo, else the working tree's repository.
`repo`	No	owner/name; defaults to the working tree's repository (its origin). Where the board lookup starts, the owner of a bare board number, and the repository of a bare issue or pull request number.
`owner`	No	A user or organization login, or "@me": the owner of a numbered board, else the only owner searched for a title.
`fields`	No	Field values to set, by field name: {"Status": "Todo", "Target": "2026-12-01"}. Values are text, a number, a YYYY-MM-DD date, an option name or an iteration title (@current is the one containing today). A column that mirrors an issue field is set with the issue tools' issue_fields, not here.
`project`	No	The board, as its number, owner/number or its URL (used as it is; a bare number takes owner, else the owner of repo or of the working tree's repository), or as its title, looked up among open boards (whole title, case-insensitive). Default: the title Roadmap.

Output Schema

Name	Required	Description
`errors`	Yes	Failed steps, keyed by step (fields.<Field name>, blocked_by, kind, labels, milestone); {} when none.
`fields`	Yes	The values that were set, by field name as the board spells it: an option or iteration by its name, a number as a number.
`content`	Yes
`created`	Yes	false when the issue or pull request was already on the board.
`item_id`	Yes	The project item's node ID (PVTI_…).
`project`	Yes
`archived`	Yes	Whether the item is archived; an archived item that was already there stays so.

Schema Changelog

Changes observed during successful MCP inspections.

Addedv0.7.0Nov 6, 2026

TDQS

A4.2/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare idempotentHint, destructiveHint and openWorldHint, and the description still adds substantial behavior: default Status resolution (Waiting vs Todo vs no default), the rule that a failed field set is reported in errors as fields.<Field name> without failing the call, blocked_by error reporting, case-insensitive name matching, and pre-write validation of every field and value. This is unusually rich disclosure of side effects and failure modes.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Almost every sentence carries real information, but the content is packed into a single dense paragraph of semicolon-chained clauses with no headings or ordering. Key facts (idempotency, default Status, error shape, auth) are buried mid-paragraph rather than front-loaded, which hurts scanability for an agent.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a five-parameter mutation with nested field objects, the description covers resolution rules, default values, partial-failure semantics, type coercion, and auth requirements. Since an output schema exists, it does not need to explain return values, yet it still clarifies the created: false flag and the errors map, leaving nothing material an agent would need before calling.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3; the description earns above that by adding the project lookup precedence (owner's boards, then repo-linked, then the organization's, then the viewer's) and the rule that a bare number resolves against repo or the working tree's origin, which the schema only partially states. Field value formats (text, number, YYYY-MM-DD, option name, iteration title, @current in UTC) are restated from the schema rather than added to.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence names a specific verb and resource: put an issue or pull request on a project board and set field values in the same call. It also implicitly distinguishes itself from github_project_item_create by stressing that adding is idempotent and returns created: false for existing items. However, it never names the sibling tools it overlaps with (github_project_item_create, github_project_item_edit) directly, so an agent must infer the boundary.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives clear routing rules: GitHub's own Title/Assignees/Labels/Milestone are set on the issue, github_issue_edit handles assignees, labels and milestone, and mirrored columns go through issue_fields rather than this tool. It also states the auth prerequisite (project scope, gh auth refresh -s project). It lacks an explicit negative case naming github_project_item_create/edit, so the guidance is clear but not exhaustive.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.