chore: add citation metadata and bump to 0.6.1 (#125)

* chore: add citation metadata and bump to 0.6.1

Adds CITATION.cff and .zenodo.json so the project can be cited and so
releases archived by Zenodo carry a DOI attributed to an ORCID iD rather
than to a bare GitHub username.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* style: format CITATION.cff to satisfy prettier

Prettier parses .cff as YAML and the file failed format:check.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>

Author: AndrewAvery7

.zenodo.json

@@ -0,0 +1,22 @@
+{
+  "title": "PromptSpend: an LLM cost estimator over a self-verifying pricing catalog",
+  "description": "<p>A web application and open dataset for estimating, comparing and understanding what large language model API usage will cost.</p><p>Prices are re-checked every morning by an automated pipeline that reconciles two independent upstream sources against hand-verified vendor overrides. Every price carries its source URL and the date it was last confirmed, disagreements between sources are surfaced for review rather than silently resolved, and every change to the catalog lands as a reviewable commit &mdash; so the provenance of any figure can be reconstructed from the repository history.</p><p>The release ships the web application, a public keyless pricing API, an MCP (Model Context Protocol) server for coding agents, and a VS Code extension that shows model prices and their last-confirmed date inline in the editor.</p>",
+  "upload_type": "software",
+  "access_right": "open",
+  "license": "MIT",
+  "creators": [
+    {
+      "name": "Avery, Andrew",
+      "orcid": "0009-0005-5476-1213"
+    }
+  ],
+  "keywords": [
+    "large language models",
+    "API pricing",
+    "cost estimation",
+    "tokenization",
+    "open dataset",
+    "data provenance",
+    "reproducibility"
+  ]
+}

CHANGELOG.md

@@ -11,6 +11,15 @@ Work that landed after `v0.6.0` was tagged. The tag points at `fd54a49` and
 these commits came after it, so they are recorded here rather than folded into
 a release that does not contain them.
 
+### Added
+
+- **`CITATION.cff` and `.zenodo.json`.** The project can now be cited. GitHub
+  renders a "Cite this repository" button from the first; the second tells
+  Zenodo how to describe each release it archives, so releases receive a DOI
+  attributed to an ORCID iD rather than to a bare GitHub username. The pricing
+  catalog is deposited separately as a dated dataset, so that published cost
+  estimates can cite the price basis they actually relied on.
+
 ### Fixed
 
 - **The results panel announced a price change no vendor had made.** The chip

CITATION.cff

@@ -0,0 +1,28 @@
+cff-version: 1.2.0
+message: 'If you use PromptSpend or its pricing catalog in your work, please cite it as below.'
+type: software
+title: 'PromptSpend'
+abstract: >-
+  A web application and open dataset for estimating, comparing and understanding
+  what large language model API usage will cost. Prices are re-checked every
+  morning by an automated pipeline that reconciles two independent upstream
+  sources against hand-verified vendor overrides; every price carries its source
+  URL and the date it was last confirmed, and every change lands as a reviewable
+  commit. Ships with a public keyless pricing API, an MCP server, and a VS Code
+  extension.
+authors:
+  - family-names: Avery
+    given-names: Andrew
+    orcid: 'https://orcid.org/0009-0005-5476-1213'
+repository-code: 'https://github.com/AndrewAvery7/promptspend'
+url: 'https://promptspend.com'
+license: MIT
+version: 0.6.1
+date-released: '2026-08-30'
+keywords:
+  - large language models
+  - api pricing
+  - cost estimation
+  - tokenization
+  - open dataset
+  - reproducibility

package.json

@@ -1,6 +1,6 @@
 {
   "name": "promptspend",
-  "version": "0.6.0",
+  "version": "0.6.1",
   "private": false,
   "description": "Estimate, compare and understand LLM API costs — with pricing that syncs itself daily.",
   "license": "MIT",