post-to-discussion

name	post-to-discussion
description	Share your completed meta-analysis project to the htlin222/meta-pipe GitHub Discussions board. Use when a project reaches a milestone (analysis complete, manuscript drafted) and the user wants to showcase their work.

Post to Discussion

Overview

Share your meta-analysis project as a GitHub Discussion on htlin222/meta-pipe. This lets other users see real examples of completed analyses, learn from different clinical questions, and reuse methodological approaches.

When to Use

• User says "share", "post", "publish to discussion", or "show and tell"
• A project has completed at least through Stage 06 (analysis)
• User wants to showcase their work to the meta-pipe community

Prerequisites

1. GitHub CLI (gh) must be authenticated: gh auth status
2. Project figures committed and pushed — PNGs are gitignored by default, so use git add -f for figures
3. Project must be on a pushed branch so image URLs resolve via raw.githubusercontent.com

Workflow

Step 1: Ensure figures are committed and pushed

# Force-add PNGs (they're gitignored globally)
git add -f projects/<project-name>/06_analysis/figures/*.png
git add -f projects/<project-name>/06_analysis/tables/*.png
git add -f projects/<project-name>/07_manuscript/figures/*.png

# Commit and push
git commit -m "Add <project-name> figures for discussion post"
git push origin <branch>

Step 2: Get the image base URL

COMMIT_SHA=$(git rev-parse HEAD)
IMG_BASE="https://raw.githubusercontent.com/htlin222/meta-pipe/${COMMIT_SHA}/projects/<project-name>"

Step 3: Get the Discussion category ID

The repo has these categories:

Category	ID	Use for
Show and tell	DIC_kwDORJgSmM4C4-A3	Completed projects (default)
Ideas	DIC_kwDORJgSmM4C4-A2	Project proposals
General	DIC_kwDORJgSmM4C4-A0	Questions, updates
Q&A	DIC_kwDORJgSmM4C4-A1	Help requests

Repository ID: R_kgDORJgSmA

Step 4: Create the discussion via GraphQL

gh api graphql -f query='
mutation {
  createDiscussion(input: {
    repositoryId: "R_kgDORJgSmA",
    categoryId: "DIC_kwDORJgSmM4C4-A3",
    title: "<Your project title>",
    body: "<Markdown body with images>"
  }) {
    discussion {
      id
      number
      url
    }
  }
}'

Step 5: Update the body (if needed)

gh api graphql -f query='
mutation {
  updateDiscussion(input: {
    discussionId: "<discussion_id>",
    body: "<updated body>"
  }) {
    discussion { url }
  }
}'

Body Template

Use this Markdown template for the discussion body. Adapt sections based on your project type (pairwise MA, NMA, pooled proportion).

## 🔬 <Project Title>

**<One-line summary of the research question and key finding.>**

> Project path: `projects/<project-name>/`

---

### Background

<2-3 sentences: clinical context, why this question matters, gap in literature.>

### Included Studies

| Study | Intervention | N | Primary Outcome |
|-------|-------------|---|-----------------|
| ... | ... | ... | ... |

---

### Figure 1: <Title>

![Figure 1](<IMG_BASE>/07_manuscript/figures/Figure1_xxx.png)

<1-2 sentence caption.>

---

### Figure 2: <Title>

![Figure 2](<IMG_BASE>/07_manuscript/figures/Figure2_xxx.png)

<1-2 sentence caption.>

---

### Key Findings

1. **Finding 1** — ...
2. **Finding 2** — ...
3. **Finding 3** — ...

### Clinical Implications

<Brief actionable summary for clinicians.>

### Methodology

- **Analysis**: <Bayesian NMA / Pairwise MA / Pooled proportion>
- **Software**: R (gemtc/netmeta/meta/metafor)
- **Heterogeneity**: I² = X%
- **Quality**: <SUCRA / GRADE / RoB summary>

### Pipeline

<Which stages were completed, total files, key artifacts.>

---

*Generated with [meta-pipe](https://github.com/htlin222/meta-pipe) — AI-assisted meta-analysis pipeline*

Tips

• Image sizing: GitHub renders full-width by default. Use <img src="..." width="600"> for smaller figures.
• Large bodies: Use a shell variable with heredoc and pipe through jq -Rs . for proper JSON escaping.
• Multiple figures: Include 3-5 key figures max. Link to the repo for supplementary materials.
• Tags: GitHub Discussions don't support tags, but you can add hashtags in the body (e.g., #NMA #DLBCL #oncology).

Example

See Discussion #30: Frontline DLBCL NMA for a complete example with 5 figures, results tables, and clinical recommendations.

Pipeline Navigation

Step	Skill	Stage
Prev	/ma-peer-review	08-09 QA & Review
This	/post-to-discussion	Share your work
All	/ma-end-to-end	Full pipeline