Add R tool to build Gmail newsletter digests with local Ollama and README

[charles-hebert]

Motivation

• Provide a lightweight, local workflow to pull newsletter-like emails from Gmail and generate a concise digest using a local Ollama model.
• Enable user preference guidance and a recommended feedback loop to iteratively improve relevance while keeping data local/private.

Description

• Add gmail_newsletter_digest.R, a CLI R script that fetches messages via gmailr, extracts text, builds a JSON prompt, and calls a local Ollama instance via httr2 to generate a digest.
• Implement helper functions including fetch_newsletters, extract_text, build_prompt, and call_ollama, and add argument parsing with argparse and default flags like --query, --ollama-model, and --output.
• Add NEWSLETTER_DIGEST_README.md with setup instructions, example Rscript run command, recommended Gmail queries, and a simple preference/rating feedback loop for matching improvements.
• Document required packages and note use of credentials.json for Gmail OAuth and default Ollama host http://127.0.0.1:11434.

Testing

• No automated tests were run for this change.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5b20d36128

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Iterate with gm_id instead of ids$id

gm_messages() returns a gmail_messages collection, not a flat object with an id field, so ids$id is NULL and this loop never runs. In practice, fetch_newsletters() returns an empty list even when the query matches emails, which makes main() always emit "No messages found for this query." Extract message IDs first (e.g., via gm_id(ids, "message_id")) before calling gm_message().

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Use supported base64 URL decoder

extract_text() calls base64urldecode(), which is not the current gmailr decode helper name, so decoding message bodies can fail at runtime with a missing-function error. Once any fetched message contains body data, digest generation breaks instead of producing output; switch to the supported decoder (base64url_decode/base64url_decode_to_char).

Useful? React with 👍 / 👎.