# Developer Protocol

**Server:** noaa-climate-mcp-server
**Version:** 0.9.0
**Framework:** [@cyanheads/mcp-ts-core](https://www.npmjs.com/package/@cyanheads/mcp-ts-core) `^0.13.22`
**Engines:** Bun ≥1.4.4, Node ≥24.0.0
**MCP SDK:** `@modelcontextprotocol/server` ^2.4.0
**Zod:** ^4.6.7

> **Read the framework docs first:** `node_modules/@cyanheads/mcp-ts-core/CLAUDE.md` contains the full API reference — builders, Context, error codes, exports, patterns. This file covers server-specific conventions only.

---

## Domain

NOAA Climate Data Online (CDO) API v2 — historical weather data — plus three separate NCEI sources: the 1991–2020 U.S. Climate Normals (Access Data Service) and two bulk-CSV corpora, the Storm Events Database and Billion-Dollar Weather and Climate Disasters. 11 tools, 2 data resources, and 1 published agent skill:

- `noaa_climate_list_datasets` — list available CDO datasets
- `noaa_climate_list_data_categories` — list measurement category groups
- `noaa_climate_list_data_types` — list measurement labels by dataset or category
- `noaa_climate_list_location_categories` — list the location categories that scope location search
- `noaa_climate_find_locations` — search geographic locations by category, with an optional client-side name filter
- `noaa_climate_find_stations` — search weather stations by location, bounding box, dataset
- `noaa_climate_get_station` — fetch full metadata for a single station
- `noaa_climate_fetch_data` — fetch historical observations with date range validation and unit selection
- `noaa_climate_get_normals` — read one station's 1991–2020 U.S. Climate Normals (monthly, daily, hourly, annual/seasonal) in NCEI's native units, with completeness flags; CDO's `NORMAL_*` datasets are the superseded 1981–2010 normals
- `noaa_climate_search_storm_events` — search NCEI Storm Events for one year (severe-weather events, damage, casualties, narratives; county rows carry `cdoLocationId`, and `countyFips` selects one county)
- `noaa_climate_get_billion_dollar_disasters` — query NCEI’s retired 1980–2024 Billion-Dollar Weather and Climate Disasters archive, per disaster or per year. Every successful response carries a retirement notice composed with paging guidance. CPI adjustments stay as archived, without current-price rebasing or an inferred base year; `coveredYears` stays row-derived. State shape is advertised as `^\s*[A-Za-z]{2}\s*$`; well-formed unavailable codes remain `unknown_state`.
- `noaa://datasets` resource — all CDO datasets as injectable context
- `noaa://stations/{stationId}` resource — station metadata by ID

CDO station inputs share `stationIdFilter` / `stationIdArrayFilter` and `normalizeStationId` in `shared/validation.ts`: trim, uppercase, and prefix bare 11-character GHCN aliases; generic IDs remain opaque. `noaa_climate_get_normals` is the exception: `ghcnStationIdFilter` / `normalizeGhcnStationId` accept only a GHCN-Daily ID and send the bare form, uppercasing only an ID written all in lowercase because NCEI matches case-sensitively. Successful tools compose normalization notices; the station resource keeps its existing result shape. `/data` alone checks the actual encoded query against 4000 characters before retry/fetch. Fetch-data keeps raw CDO types distinct from its output: documented PRECIP_HLY/HPCP and PRECIP_15/QPCP missing sentinels omit `value` and set `missing: true`; only GHCND attributes gain decoded `flags`. Missing/QC counts are page-local, and raw-unit guidance is datatype-qualified. All notices compose with clamp and paging guidance.

**Services:**

- `src/services/cdo/cdo-service.ts` — HTTP client with retry/backoff, `NOAA_CDO_TOKEN` header auth.
- `src/services/storm-events/storm-events-service.ts` — NCEI bulk CSV. No token, no JSON envelope, no `resolveCollectionTotal()`. Resolves each year's `_c<publishDate>`-suffixed filename from the live directory index (it is not constructible), decompresses explicitly (the files carry no `Content-Encoding`), and streams a year through the shared CSV reader so the decompressed ~70 MB is never held. Caches the listing and two years of compressed bytes for six hours. Damage parsing lives in `damage.ts` — an empty cell is "not reported", never `0`. `countyCodes()` pads `STATE_FIPS`/`CZ_FIPS` and composes `cdoLocationId` only for `C` rows with a state code in 01–56 other than Alaska's 02; the projection and the `countyFips` filter both read it.
- `src/services/normals/normals-service.ts` — NCEI Access Data Service (`/access/services/data/v1`) for 1991–2020 normals. No token. One dataset per granularity, every datatype named explicitly (the service's own monthly default omits `MLY-TAVG-NORMAL`), `units` never sent, rows of left-padded strings parsed into one record per datatype, a bounded timeout and a total deadline around `withRetry`. A datatype outside the verified `UNITS` table gets no `unit`.
- `src/services/billion-dollar-disasters/billion-dollar-disasters-service.ts` — NCEI Billion-Dollar Disasters. No token, four small static CSVs (per-event and per-year, national and per-state), each opening with preamble lines ahead of the header. **Each file declares its own cost unit and they differ** — millions for per-event, billions for the national per-year, millions again for a per-state per-year — so the unit is parsed from the file's preamble and applied at one site, `toUsd()`, converting everything to whole US dollars. An unrecognized or absent unit is `malformed_export`, never a guess.
- `src/services/csv/csv-stream-reader.ts` — the RFC 4180 reader both bulk-CSV services parse with.

---

## What's Next?

When the user asks what's next or needs direction, suggest options based on the current project state. Common next steps:

1. **Re-run the `setup` skill** — ensures CLAUDE.md, skills, structure, and metadata are populated and up to date with the current codebase
2. **Run the `design-mcp-server` skill** — if the tool/resource surface hasn't been mapped yet, work through domain design
3. **Add tools/resources/prompts** — scaffold new definitions using the `add-tool`, `add-app-tool`, `add-resource`, `add-prompt` skills
4. **Add services** — scaffold domain service integrations using the `add-service` skill
5. **Add tests** — scaffold tests for existing definitions using the `add-test` skill
6. **Field-test definitions** — exercise tools/resources/prompts with real inputs using the `field-test` skill, get a report of issues and pain points
7. **Run `devcheck`** — lint, format, typecheck, and security audit
8. **Run the `security-pass` skill** — audit handlers for MCP-specific security gaps: output injection, scope blast radius, input sinks, tenant isolation
9. **Run the `polish-docs-meta` skill** — finalize README, CHANGELOG, metadata, and agent protocol for shipping
10. **Run the `maintenance` skill** — investigate changelogs, adopt upstream changes, and sync skills after `bun update --latest`

Tailor suggestions to what's actually missing or stale — don't recite the full list every time.

---

## Core Rules

- **Logic throws, framework catches.** Tool/resource handlers are pure — throw on failure, no `try/catch`. Plain `Error` is fine; the framework catches, classifies, and formats. Use error factories (`notFound()`, `validationError()`, etc.) when the error code matters.
- **Use `ctx.log`** for request-scoped logging. No `console` calls.
- **Use `ctx.state`** for tenant-scoped storage. Never access persistence directly.
- **Need extra input after a call?** Return `ctx.requestInput(...)` and read `ctx.inputs` on the next call. This server has no current elicitation call sites.
- **Secrets in env vars only** — never hardcoded.
- **Cut noise.** Add only what earns its place: no speculative generality, no guards for states the framework already prevents (Zod-validated params, classified errors), no abstraction until a third caller proves it, no option nothing sets.
- **Close the loop on issues.** When implementing work tracked by a GitHub issue, comment on the issue with what landed and close it. Do both — a comment without a close leaves stale issues open; a close without a comment leaves no record of what shipped. The comment is for future readers — state the concrete changes, not the conversation that produced them.

---

## Patterns

The implementation itself is the canonical example. Start with these representative definitions:

| Primitive | Source | Pattern demonstrated |
|:---|:---|:---|
| Tool | `src/mcp-server/tools/definitions/noaa-climate-list-location-categories.tool.ts` | Strict input/output schemas, declared errors, success enrichment, and equivalent `format()` text |
| Tool | `src/mcp-server/tools/definitions/noaa-climate-get-station.tool.ts` | Cross-tool ID workflow, typed failure routing, and sparse upstream fields preserved as optional |
| Resource | `src/mcp-server/resources/definitions/noaa-station.resource.ts` | URI-template params, request context propagation, and typed not-found recovery |
| Config | `src/config/server-config.ts` | Lazy `parseEnvConfig()` mapping from `token` to the required `NOAA_CDO_TOKEN` env var |
| Entry point | `src/index.ts` | Machine-name identity, tool/resource registration, service initialization, and server instructions |

There are no prompts. Do not add prompt examples or prompt directories to documentation until the server exposes a real prompt.

`createApp()` identity is deliberately the unscoped machine name on both fields:

```ts
await createApp({
  name: 'noaa-climate-mcp-server',
  title: 'noaa-climate-mcp-server',
  sessionMode: 'stateless',
  // tools, resources, instructions, and setup are registered in src/index.ts
});
```

Do not duplicate `package.json` description or homepage metadata into `createApp()`; the framework derives canonical metadata from the package.

`sessionMode: 'stateless'` declares the HTTP session posture in `src/` rather than leaving it to a deployment's `MCP_SESSION_MODE`, which still wins whenever it carries a meaningful value. Stateless is correct here: no tool asks the caller for input mid-handler via `ctx.requestInput`, so nothing needs `require: 'stateful'`. No service allocates a watcher, socket, or ref'd timer — the three caches are `Date.now()`-stamped in-memory maps — so `createApp({ teardown })` has nothing to release and is not declared.

---

## Context

Handlers receive a unified `ctx` object. Key properties:

| Property | Description |
|:---------|:------------|
| `ctx.log` | Request-scoped dual-sink logger — server output plus MCP `notifications/message` when the client enables that level. Never log the CDO token. |
| `ctx.enrich` | Success-path agent context (empty-result notices, query echo, pagination totals) — `ctx.enrich(...)` or `.notice()` / `.total()` / `.echo()` / `.truncated()`. Reaches `structuredContent` and `content[]`; lands only when the definition declares an `enrichment` block (no-op otherwise). |
| `ctx.fail` | Emit a declared tool/resource failure; the framework fills the entry's recovery hint on the wire. `ctx.recoveryFor(reason)` reads a declared hint, for forwarding one reason's hint onto another. |
| `ctx.signal` | `AbortSignal` for cancellation. |
| `ctx.requestId` / `ctx.tenantId` / `ctx.auth` | Propagated by passing the complete `ctx` as `parentContext` when a service creates an operation context. `ctx.requestId` is the ID every log record of the call carries and its error envelope returns as `data.requestId`. |

---

## Errors

Handlers throw — the framework catches, classifies, and formats.

**Recommended: typed error contract.** Declare `errors: [{ reason, code, when, recovery, retryable?, severity?, thrownBy? }]` on `tool()` / `resource()` to receive `ctx.fail(reason, …)` typed against the reason union. TypeScript catches typos at compile time, `data.reason` is auto-populated for observability, linter enforces conformance against the handler body. `recovery` is required (≥ 5 words, lint-validated) — the single source of truth for the agent's next move. The framework puts it on the wire whenever a failure carrying that `reason` arrives without a hint — a bare `ctx.fail('reason')` or a service throw with `data: { reason }` — as `data.recovery.hint`, mirrored into `content[]` text unless the message already contains it verbatim; override with an explicit `{ recovery: { hint: '...' } }` when dynamic runtime context matters. Every error envelope also carries `data.requestId`, the id the server's log records for that call carry, and `content[]` closes with `(reason … · request <id>)`. Mark an entry the service layer throws with `thrownBy: 'service'` so `error-contract-unthrown` skips it — lint-only metadata, nothing at runtime reads it. Baseline codes (`InternalError`, `ServiceUnavailable`, `Timeout`, `ValidationError`, `SerializationError`, `RequestCancelled`) bubble freely and don't need declaring.

```ts
import { JsonRpcErrorCode } from '@cyanheads/mcp-ts-core/errors';

errors: [
  { reason: 'no_match', code: JsonRpcErrorCode.NotFound,
    when: 'No item matched the query',
    recovery: 'Broaden the query or check the spelling and try again.' },
],
async handler(input, ctx) {
  const item = await db.find(input.id);
  if (!item) throw ctx.fail('no_match', `No item ${input.id}`);
  return item;
}
```

**Declare contracts inline on each tool.** The contract is part of the tool's public surface — one file should give the full picture. Don't extract a shared `errors[]` constant; per-tool repetition is the intended cost of locality.

**Fallback (no contract entry fits):** throw via factories or plain `Error`.

```ts
// Error factories — explicit code
import { notFound, serviceUnavailable } from '@cyanheads/mcp-ts-core/errors';
throw notFound('Item not found', { itemId });
throw serviceUnavailable('API unavailable', { url }, { cause: err });

// Plain Error — framework auto-classifies from message patterns
throw new Error('Item not found');           // → NotFound
throw new Error('Invalid query format');     // → ValidationError

// McpError — when no factory exists for the code
import { McpError, JsonRpcErrorCode } from '@cyanheads/mcp-ts-core/errors';
throw new McpError(JsonRpcErrorCode.InitializationFailed, 'Connection failed', { pool: 'primary' });
```

See framework CLAUDE.md and the `api-errors` skill for the full auto-classification table, all available factories, and the contract reference.

---

## Structure

```text
src/
  index.ts                              # createApp() entry point
  config/
    server-config.ts                    # Required NOAA_CDO_TOKEN (Zod schema)
  services/
    cdo/                                # CDO API client, pagination, raw types
    csv/                                # Shared streaming RFC 4180 reader
    normals/                            # 1991–2020 normals: Access Data Service client, unit labels
    storm-events/                       # Storm Events listing, download, parsing
    billion-dollar-disasters/           # Disaster exports, unit normalization
  mcp-server/
    tools/definitions/
      noaa-climate-*.tool.ts            # 11 tool definitions
      shared/                           # Shared validation and upstream routing
    resources/definitions/
      noaa-datasets.resource.ts         # Dataset catalog resource
      noaa-station.resource.ts          # Station URI-template resource
```

---

## Naming

| What | Convention | Example |
|:-----|:-----------|:--------|
| Files | kebab-case with suffix | `search-docs.tool.ts` |
| Tool/resource/prompt names | snake_case | `search_docs` |
| Directories | kebab-case | `src/services/doc-search/` |
| Descriptions | Single string or template literal, no `+` concatenation | `'Search items by query and filter.'` |

---

## Skills

`skills/noaa-climate-guide/SKILL.md` is the published caller guide for climate comparisons, coverage, observation quality, and storm-to-station investigation. `src/index.ts` serves it through the MCP Skills extension and a skill resource alongside the two data resources. Plugin, npm, Docker, and MCPB delivery preserve the root `skills/` tree.

Skills are modular instructions in `framework-skills/` at the project root. Read them directly when a task matches — e.g., `framework-skills/add-tool/SKILL.md` when adding a tool. `bun run list-skills` prints the full registry. The directory is deliberately not `skills/`: Claude Code and Codex auto-load a plugin's root `skills/`, so a server that ships `.claude-plugin/` or `.codex-plugin/` would hand these development skills to every agent that installs it. Keep `skills/` free for skills meant for those agents.

**Agent skill directory:** Copy skills into the directory your agent discovers (Claude Code: `.claude/skills/`, others: equivalent). Skills then load as context without referencing `framework-skills/` paths. After framework updates, run the `maintenance` skill — Phase B re-syncs the agent directory.

Available skills:

| Skill | Purpose |
|:------|:--------|
| `setup` | Post-init project orientation |
| `design-mcp-server` | Design tool surface, resources, and services for a new server |
| `add-tool` | Scaffold a new tool definition |
| `add-skill` | Author and publish agent-facing skills from root `skills/`, served over MCP |
| `add-app-tool` | Scaffold an MCP App tool + paired UI resource |
| `add-resource` | Scaffold a new resource definition |
| `add-prompt` | Scaffold a new prompt definition |
| `add-service` | Scaffold a new service integration |
| `add-test` | Scaffold test file for a tool, resource, or service |
| `field-test` | Exercise tools/resources/prompts with real inputs, verify behavior, report issues |
| `tool-defs-analysis` | Read-only audit of MCP definition language across the surface — voice, leaks, defaults, recovery hints, output descriptions |
| `security-pass` | Audit server for MCP-flavored security gaps: output injection, scope blast radius, input sinks, tenant isolation |
| `code-simplifier` | Post-session cleanup against `git diff` — modernize syntax, consolidate duplication, align with the codebase |
| `polish-docs-meta` | Finalize docs, README, metadata, and agent protocol for shipping |
| `git-wrapup` | Land working-tree changes as a commit stack — version bump, changelog, verify, commit by concern, release commit on top. No tag, no push to main; opens the release PR when the project declares release PR mode |
| `release-pr-review` | Review pass on an open release PR — simplifier + correctness review, fixes as ordinary commits on top of the stack, PR body kept in sync. Release PR mode only |
| `release-and-publish` | Fast-forward merge (release PR mode) + tag + push + npm + MCP Registry + GH Release + Docker. Picks up from `git-wrapup` |
| `maintenance` | Investigate changelogs, adopt upstream changes, sync skills to agent dirs |
| `orchestrations` | Chain task skills into a gated multi-phase pipeline — build-out, QA-fix, update-ship — when you can spawn sub-agents |
| `report-issue-framework` | File a bug or feature request against `@cyanheads/mcp-ts-core` via `gh` CLI |
| `report-issue-local` | File a bug or feature request against this server's own repo via `gh` CLI |
| `techniques` | Catalog of response/data-shaping techniques — overflow handling, payload shaping, retrieval patterns |
| `api-auth` | Auth modes, scopes, JWT/OAuth |
| `api-canvas` | DataCanvas: register tabular data, run SQL, export, plus the `spillover()` helper for big result sets — Tier 3 opt-in |
| `api-config` | AppConfig, parseConfig, env vars |
| `api-context` | Context interface, RequestContext, logger, state, multi-round-trip input |
| `api-errors` | McpError, JsonRpcErrorCode, error patterns |
| `api-linter` | Definition linter rule catalog — invoked by `bun run lint:mcp` and `devcheck` |
| `api-mirror` | MirrorService: persistent self-refreshing local mirror (embedded SQLite + FTS5) of a bulk upstream dataset — Tier 3 opt-in |
| `api-services` | LLM, Speech, Graph services |
| `api-testing` | createMockContext, test patterns |
| `api-utils` | Formatting, parsing, security, pagination, scheduling, telemetry helpers |
| `api-telemetry` | OTel catalog: spans, metrics, completion logs, env config, cardinality rules |
| `api-workers` | Cloudflare Workers runtime |

**Chaining skills into pipelines.** When the user wants a multi-phase effort — build this server out, QA-and-fix the surface, update-and-ship — *and you can spawn sub-agents*, `framework-skills/orchestrations/SKILL.md` sequences the task skills above into a gated pipeline with verification at each step. Read it to drive the run. Optional: skip it if you can't orchestrate sub-agents, and ignore it entirely if you were *spawned* as one — you've already been scoped to a single phase.

When you complete a skill's checklist, check the boxes and add a completion timestamp at the end (e.g., `Completed: 2026-03-11`).

---

## Commands

**Runtime:** Scripts use Bun's native TypeScript execution — `bun run <cmd>` is the standard invocation. `npm run <cmd>` also works (npm delegates to bun).

| Command | Purpose |
|:--------|:--------|
| `bun run build` | Compile TypeScript |
| `bun run rebuild` | Clean + build |
| `bun run clean` | Remove build artifacts |
| `bun run devcheck` | Lint + format + typecheck + security + changelog sync |
| `bun run audit:fix` | `bun audit fix` — upgrade vulnerable packages to the lowest safe version within existing ranges (`--dry-run` previews, `--latest` rewrites ranges). First response when `devcheck` flags a transitive advisory; then `bun update <name>`, then `bun dedupe` |
| `bun run audit:refresh` | Delete `bun.lock` and reinstall. Last resort after `audit:fix`, `bun update <name>`, and `bun dedupe` — re-resolves every ranged dep (the framework pin included) and rewrites the lockfile as `lockfileVersion: 2` |
| `bun run lint:mcp` | Run the MCP definition linter standalone (rule catalog: `api-linter` skill) |
| `bun run lint:packaging` | Packaging surface checks — `server.json`/`manifest.json` env-var parity, `.mcpbignore` coverage, plugin manifests, the published `skills/` tree (run by devcheck) |
| `bun run list-skills` | Print the skill registry |
| `bun run tree` | Generate directory structure doc |
| `bun run format` | Auto-fix formatting (safe fixes only) |
| `bun run format:unsafe` | Also apply Biome's unsafe autofixes — review the diff; they can change behavior |
| `bun run test` | Run the hermetic `unit` project (Vitest — use `bun run test`, not `bun test`) |
| `bun run test:live` | Opt-in `live` project — resolves every example identifier in tool descriptions, README, and `docs/design.md` against the live CDO API. Needs `NOAA_CDO_TOKEN`; kept out of `test` so the default lane stays offline |
| `bun run start:stdio` | Production mode (stdio) |
| `bun run start:http` | Production mode (HTTP) |
| `bun run changelog:build` | Regenerate `CHANGELOG.md` from `changelog/*.md` |
| `bun run changelog:check` | Verify `CHANGELOG.md` is in sync (used by devcheck) |
| `bun run bundle` | Build, pack, and clean a `.mcpb` for one-click Claude Desktop install |

**CI is one file.** `.github/workflows/codeql.yml` (scaffolded) is the only GitHub Actions workflow: CodeQL is GitHub-owned end to end, and the file runs only while the repo's CodeQL *default setup* is turned off. Verification — `devcheck`, tests, the release gates — runs locally; don't add a workflow that re-runs it.

---

## Bundling

`npm run bundle` produces a `.mcpb` extension bundle for one-click install in Claude Desktop. `mcpb pack` reads `.mcpbignore`, not `.gitignore`, so every root path the bundle must not ship — source, tests, docs, caches, logs, agent docs, plugin manifests, gitignored local files — needs a root-anchored line there (`/docs/`, never `docs/`, which also strips `node_modules/*/docs/`); `lint:packaging` names each missing or unanchored line. Root `skills/` ships on purpose. The pack step is followed by `scripts/clean-mcpb.ts`, which prunes dev dependencies (`mcpb clean`) and strips four classes of `node_modules/**` content that root-anchored `.mcpbignore` patterns cannot reach: dependency-shipped agent docs — `framework-skills/`, `skills/`, `.claude/`, `.agents/`, `AGENTS.md`, and `CLAUDE.md` at a package's root, and `SKILL.md` anywhere (a `skills/` deeper in a package is runtime code and stays); platform-specific native bindings, which would otherwise lock the bundle to the platform it was packed on; node-gyp build output beside a `binding.gyp` (`config.gypi`, Makefiles, `obj.target/`, `.deps/`, object files and debug symbols), which records the packing machine's paths — compiled `*.node` addons stay; and build-only package files — the framework's `changelog/`, `scripts/`, `templates/`, and config bases (`biome.json`, `vitest.config.base.mjs`, `config/tsconfig.base.json`), and `better-sqlite3`'s `deps/` and `src/` C sources. A server using DataCanvas therefore ships a portable bundle without the DuckDB native — `@duckdb/node-api` is an optional peer loaded lazily, so canvas tools report an actionable install hint and every other tool works normally. MCPB is stdio-only — HTTP and Cloudflare Workers deployments are unaffected. Consumers who don't need it can delete `manifest.json` and `.mcpbignore`; `lint:packaging` skips cleanly.

**Adding an env var requires both files:** `server.json` (registry discovery, `environmentVariables[]`) and `manifest.json` (bundle install UX, `mcp_config.env` + `user_config`). `lint:packaging` (run by `devcheck`) verifies the env var names match, that every `user_config` option is wired into `mcp_config.env` as `"X": "${user_config.X}"` (the host substitutes nothing else — `"${X}"` reaches the server as that literal string), and that an optional string option carries `"default": ""`.

**README install badges** (Claude Desktop `.mcpb`, Cursor, VS Code) and the `base64` / `encodeURIComponent` config-generation commands are ship-time concerns — run the `polish-docs-meta` skill, which carries the badge format, layout, and generation snippets in `framework-skills/polish-docs-meta/references/readme.md`.

---

## Changelog

Directory-based, grouped by minor series via the `.x` semver-wildcard convention. Source of truth: `changelog/<major.minor>.x/<version>.md` (e.g. `changelog/0.1.x/0.1.0.md`) — one file per release, shipped in the npm package. At release, author the per-version file with a concrete version and date, then run `npm run changelog:build` to regenerate the rollup. `changelog/template.md` is a **pristine format reference** — never edited or moved; read it for the frontmatter + section layout when scaffolding. `CHANGELOG.md` is a **navigation index** (header + link + summary per version), regenerated by `npm run changelog:build` — devcheck hard-fails on drift; never hand-edit it.

Each per-version file opens with YAML frontmatter:

```markdown
---
summary: "One-line headline, ≤350 chars"  # required — powers the rollup index
breaking: false                            # optional — true flags breaking changes
security: false                            # optional — true ONLY for a source-code security fix, never a dependency CVE bump
---

# 0.1.0 — YYYY-MM-DD
...
```

`breaking: true` renders a `· ⚠️ Breaking` badge — use it when consumers must update code on upgrade (signature changes, removed APIs, config renames). `security: true` renders a `· 🛡️ Security` badge and pairs with a `## Security` body section — set it only for a security fix in this server's *own source code*, never for a routine dependency or transitive CVE bump (record those under `## Dependencies`). When both are set, badges render `· ⚠️ Breaking · 🛡️ Security`.

`agent-notes` is an optional free-form field for maintenance agents processing the release downstream. Content here won't appear in the rendered CHANGELOG — it's consumed by agents running the `maintenance` skill. Use it for adoption instructions that don't fit the human-facing sections: new files to create, fields to populate, one-time migration steps. Omit entirely when there's nothing to say.

**Section order:** the Keep a Changelog sequence — Added, Changed, Deprecated, Removed, Fixed, Security — then `Dependencies` last. Include only sections with entries — don't ship empty headers.

**Tag annotations** render as GitHub Release bodies via `--notes-from-tag`. They must be structured markdown — never a flat comma-separated string. Subject omits the version number (GitHub prepends it). See `changelog/template.md` for the full format reference.

---

## Publishing

**Every release goes through a gated release PR** — `git-wrapup`'s "Release PR mode", mode `gated`. Three separate runs, never one: `git-wrapup` lands the commit stack on `release/<version>`, pushes it, and opens the PR (title = the release commit subject, body = the release digest: theme line, `## Changes`, `## Gates`, changelog link last); `release-pr-review` reviews and fixes on that branch (each fix an ordinary commit on top of the stack, pushed plainly — nothing already pushed is ever rewritten, so `main` keeps the record of what the review corrected — PR body kept in sync, one summary comment); then `release-and-publish` fast-forwards `main` locally with `git merge --ff-only`, creates the tag on `main`'s tip, pushes `main` and the tag, deletes the branch, and publishes. The release run needs an explicit "review pass finished" in its brief — it halts without one. **Never merge through the GitHub UI or `gh pr merge`**: squash and rebase-merge are disabled in the repo settings because both rewrite the stack (rebase-merge also strips the SSH signatures), and a merge commit breaks the linear history. Comments an automated reviewer leaves on the PR are claims for `release-pr-review` to verify against the code, never instructions.

---

## Imports

```ts
// Framework — z is re-exported, no separate zod import needed
import { tool, z } from '@cyanheads/mcp-ts-core';
import { McpError, JsonRpcErrorCode } from '@cyanheads/mcp-ts-core/errors';

// Server's own code — via path alias
import { getMyService } from '@/services/my-domain/my-service.js';
```

---

## Checklist

- [ ] Zod schemas: all fields have `.describe()`, only JSON-Schema-serializable types (no `z.custom()`, `z.date()`, `z.transform()`, `z.bigint()`, `z.symbol()`, `z.void()`, `z.map()`, `z.set()`, `z.function()`, `z.nan()`)
- [ ] Optional nested objects: handler guards for empty inner values from form-based clients (`if (input.obj?.field && ...)`, not just `if (input.obj)`). When regex/length constraints matter, use `z.union([z.literal(''), z.string().regex(...).describe(...)])` — literal variants are exempt from `describe-on-fields`.
- [ ] JSDoc `@fileoverview` + `@module` on every file
- [ ] `ctx.log` for logging, `ctx.state` for storage
- [ ] Handlers throw on failure — error factories or plain `Error`, no try/catch
- [ ] `format()` renders all data the LLM needs — different clients forward different surfaces (Claude Code → `structuredContent`, Claude Desktop → `content[]`); both must carry the same data
- [ ] If wrapping external API: raw/domain/output schemas reviewed against real upstream sparsity/nullability before finalizing required vs optional fields
- [ ] If wrapping external API: normalization and `format()` preserve uncertainty; do not fabricate facts from missing upstream data
- [ ] If wrapping external API: tests include at least one sparse payload case with omitted upstream fields
- [ ] Registered in `createApp()` arrays (directly or via barrel exports)
- [ ] Tests use `createMockContext()` from `@cyanheads/mcp-ts-core/testing`
- [ ] `plugin.json` + `mcp.json` (Agent Plugins 1.0.0 — read by Codex, Cursor, GitHub Copilot, VS Code, Kiro) populated — `plugin.json` `name` = the unscoped repo name (lowercase `a-z`, `0-9`, `-`, `.` only); `version`, `description`, `author`, `homepage`, `repository`, `license`, `keywords` from `package.json`; no other top-level field, `extensions` included. `mcp.json` holds one `"type": "stdio"` entry keyed by the unscoped name, `npx -y <full package name>`, and the same fixed `env` as `.codex-plugin/mcp.json`. Neither file carries a user-supplied variable (no `""`, no `${…}` besides `${PLUGIN_ROOT}`/`${PLUGIN_DATA}`): keys reach Codex through `.codex-plugin/mcp.json` `env_vars`, which works only while `.codex-plugin/plugin.json` keeps `"mcpServers": "./.codex-plugin/mcp.json"`
- [ ] `.codex-plugin/plugin.json` populated — `name`, `version`, `description`, `repository`, `license` from `package.json`; `interface.displayName` = the unscoped repo name (never the npm scope — `lint:packaging` enforces this); `interface.shortDescription` from `package.json` description
- [ ] `.codex-plugin/mcp.json` updated — server name key is the unscoped repo name; every user-supplied variable (API key, contact email, instance URL) is listed in `env_vars` so Codex forwards it from the user's environment. Never write `"KEY": ""` into `env` — an empty value replaces the user's exported key and is read as unset
- [ ] `.claude-plugin/plugin.json` populated — `name`, `version`, `description`, `author`, `repository`, `license`, `keywords` from `package.json`; inline `mcpServers` entry keyed by the unscoped repo name. Every user-supplied variable is declared under `userConfig` (`type`, `title`, `description`; `sensitive: true` for keys and tokens; `required: true` or `default: ""`) and referenced from `env` as `"KEY": "${user_config.<option>}"` — mirror the `user_config` block in `manifest.json`. Never write `"KEY": ""` into `env`
- [ ] Any new root dev path has a root-anchored line in `.mcpbignore`; `bun run bundle` and `lint:packaging` pass
- [ ] If the server publishes a skill in root `skills/`: `lint:packaging` passes, and `field-test` verified every gate and identifier claim in it against live tool behavior
- [ ] `npm run devcheck` passes