Pm Tasks Trello

pm-tasks-trello

Adapter for Trello within the @llodev/pm-tasks-* family. Use the core skill's extraction phases, then apply Trello formatting and optionally publish/operate via the atlassian-trello-mcp MCP server.

Routing

Phase 4 — Trello formatting

MANDATORY — READ ENTIRE FILE references/format.md before producing any Trello-specific output. Then apply anti-patterns/tools.md § Trello.

Phase 5 — MCP publish

Prerequisites: atlassian-trello-mcp configured (see references/mcp-config.md). Env vars TRELLO_API_KEY + TRELLO_TOKEN in shell.

Strict order: 5.1 config discovery → 5.2.5 resolve labels/member → 5.2 preview & approval → 5.3 publish via MCP → 5.4 error handling.

Full sequence in references/publish.md.

Attribution (opt-in)

Before calling the MCP create tool, read config.attribution. If enabled === true, prefix the comment with the commentPrefix returned by getAttribution() and append the descriptionFooter to the end of description. In autonomous mode ([autonomous] sentinel), the commentPrefix automatically becomes the autonomousCommentPrefix. See references/attribution.md in pm-tasks-core (added in v1.2.0).

Phase 5b — Autonomous

Skip 5.2 preview & approval. Apply autonomous-mode contract from skills/pm-tasks-core/references/autonomous-mode.md. Tool-specific overlay in references/autonomous.md. Audit log entries per skills/pm-tasks-core/references/audit-log-format.md.

Phase 6 — CRUD operations (existing cards)

For verbs other than task.create, jump directly to the operation. MANDATORY — READ ENTIRE FILE references/operations.md which lists verb → MCP tool mapping and <task-ref> resolution for Trello URLs/IDs. For task.comment.add, apply attribution prefix if config.attribution.enabled === true (see Phase 5 § Attribution).

Temporal handling (create-time due, start on WIP move, and the overwrite-due-and-footer close) is documented in references/operations.md § Temporal handling.

trello.task.batch-create-with-checklists (custom verb)

Batch-creates multiple cards, each with its checklists, in bounded parallel — ~10× faster than one-at-a-time on large plans. Cards route through the audited task.create path (audit log entry per card; no automatic idempotency/dedupe on retry — re-running a batch re-creates cards that already succeeded); checklists are created two-phase (all checklists, then all items) with a concurrency cap (concurrency, default 8) that bounds each phase within one card's checklist creation. Across the batch, up to concurrency cards are also created in parallel, so peak in-flight MCP calls can reach concurrency × concurrency (64 at the default) — lower concurrency for very large batches. Trello's 300 req/10s limit still applies; a 429 degrades to a per-card error envelope, not a crash. One card failing does not abort the batch. Autonomous- gateable via autonomous.allow. Headless entry: @llodev/pm-tasks-trello/adapter → trelloBatchCreateWithChecklists. This release's speedup is parallelism; idChecklistSource template cloning (exposed by the Trello MCP) is a possible future optimization for repeated checklist templates.

Phase 7 — Plan-execution mode

When the calling agent passes a plan reference (a path to a markdown plan file, a plan slug, or an explicit list of expected task titles), this skill loads .trello.json via requireConfig and uses discoverPlanTasks (from @llodev/pm-tasks-core) to triage which tasks in the plan already exist as cards in scope. The skill returns the triage report { found, missing, ambiguous } — the calling agent decides how to act on each bucket (create missing cards via Phase 4 + 5; disambiguate by picking the right ambiguous card; proceed with existing).

The skill does not assume any particular implementation strategy. It does not drive the calling agent's task loop, does not depend on any specific orchestration framework, and does not require any markup beyond what is already documented for task.create and the autonomous-mode contract. When the calling agent finishes a task and asks the skill to record progress, it invokes the standard verbs (task.move, checklist.check, task.comment.add, task.close) directly — the same path used by Phase 5b autonomous mode.

Full contract: triggers, discovery semantics, ConfigRequiredError handling, failure modes table, and the hook contract for Phase 5 are documented in skills/pm-tasks-core/references/plan-execution.md (added in v1.9.0).

Standalone fallback

If @llodev/pm-tasks-core is not installed: ask the user for minimum input (title + checklist items) and produce a paste-ready Trello card from this content alone. Quality is degraded — no scope/audience/fidelity inference. Print: "Install @llodev/pm-tasks-core for the full flow."

Config

Lookup order: <git-root>/.trello.json → ~/.config/llodev/pm-tasks/trello.json → abort with init instructions. Schema: schemas/config.json. Secrets NEVER in JSON — only env vars / keychain.

Narration language. When .trello.json sets locale, agent-authored narration — comments the agent writes, autonomous commit/PR copy, code comments — MUST use it. The plan's language still governs card title/description/checklist text. pm-tasks-core-doctor C-LANG-1 warns if unset or set to a locale with no installed i18n bundle. See pm-tasks-core § Narration language.

Init

npx @llodev/pm-tasks-trello init

See skills/pm-tasks-core/references/init-ux.md for the shared flow.

Pass --doctor to run workspace health checks before prompting:

npx @llodev/pm-tasks-trello init --doctor

Runs core checks (C-FS-1..3, C-CFG-1..4) plus Trello-specific probes (C-TRL-1..3, gated on TRELLO_API_KEY/TRELLO_TOKEN). Full check matrix in skills/pm-tasks-core/references/doctor.md.