feat: add scripts to retrieve repositories with and without Copilot custom instructions (#159)

* feat: add scripts to retrieve repositories with and without Copilot custom instructions

* docs: update README and scripts to include authentication details for Copilot instructions

Author: joshjohanning

gh-cli/README.md

@@ -1319,6 +1319,26 @@ Generates a CSV with 4 columns:
 
 This script is useful when doing migrations, to determine the kind of actions that might be needed based on the webhooks inventory.
 
+### get-repositories-with-copilot-instructions.sh
+
+Get repositories that have Copilot custom instruction files, checking for both repository-wide (`.github/copilot-instructions.md`) and path-specific (`.github/instructions/`) instructions
+
+Usage:
+
+```shell
+./get-repositories-with-copilot-instructions.sh my-org
+```
+
+### get-repositories-without-copilot-instructions.sh
+
+Get repositories that do not have any Copilot custom instruction files (neither `.github/copilot-instructions.md` nor `.github/instructions/`)
+
+Usage:
+
+```shell
+./get-repositories-without-copilot-instructions.sh my-org
+```
+
 ### get-repository-languages-for-organization.sh
 
 Get the repository language information (ie: JavaScript, Python, etc) for all repositories in an organization. Can specify how many language results to return (top X).

gh-cli/get-repositories-with-copilot-instructions.sh

@@ -0,0 +1,55 @@
+#!/bin/bash
+
+# Get repositories that have Copilot custom instruction files
+# Checks for:
+#   - .github/copilot-instructions.md (repository-wide custom instructions)
+#   - .github/instructions/ directory (path-specific custom instructions)
+#
+#
+# Authentication:
+#   Requires a GitHub CLI token with at least read:org and repo scopes
+#   Example: gh auth refresh -h github.com -s read:org,repo
+#
+# Usage:
+#   ./get-repositories-with-copilot-instructions.sh <org>
+#   If you want to see the results in a nicely formatted table, you can pipe the output to `column`:
+#   ./get-repositories-with-copilot-instructions.sh <org> | column -ts $'\t'
+
+if [ -z "$1" ]; then
+  echo "Usage: $0 <org>"
+  exit 1
+fi
+
+org=$1
+
+echo -e "Repository\tRepo-Wide\tPath-Specific Files"
+
+gh api graphql --paginate -F owner="$org" -H "X-Github-Next-Global-ID: 1" -f query='
+  query ($owner: String!, $endCursor: String = null) {
+    organization(login: $owner) {
+      repositories(
+        first: 100
+        orderBy: { field: NAME, direction: ASC }
+        after: $endCursor
+      ) {
+        totalCount
+        pageInfo {hasNextPage endCursor}
+        nodes {
+          nameWithOwner
+          repoWide: object(expression: "HEAD:.github/copilot-instructions.md") {
+            ... on Blob {
+              byteSize
+            }
+          }
+          pathSpecific: object(expression: "HEAD:.github/instructions/") {
+            ... on Tree {
+              entries {
+                path
+              }
+            }
+          }
+        }
+      }
+    }
+  }
+' --jq '.data.organization.repositories.nodes[] | select(.repoWide != null or .pathSpecific != null) | [.nameWithOwner, (if .repoWide != null then "yes" else "no" end), (if .pathSpecific != null then (.pathSpecific.entries | length | tostring) else "0" end)] | @tsv'

gh-cli/get-repositories-without-copilot-instructions.sh

@@ -0,0 +1,50 @@
+#!/bin/bash
+
+# Get repositories that do not have any Copilot custom instruction files
+# Checks for the absence of both:
+#   - .github/copilot-instructions.md (repository-wide custom instructions)
+#   - .github/instructions/ directory (path-specific custom instructions)
+#
+# Authentication:
+#   Requires a GitHub CLI token with at least read:org and repo scopes
+#   Example: gh auth refresh -h github.com -s read:org,repo
+#
+# Usage:
+#   ./get-repositories-without-copilot-instructions.sh <org>
+
+if [ -z "$1" ]; then
+  echo "Usage: $0 <org>"
+  exit 1
+fi
+
+org=$1
+
+gh api graphql --paginate -F owner="$org" -H "X-Github-Next-Global-ID: 1" -f query='
+  query ($owner: String!, $endCursor: String = null) {
+    organization(login: $owner) {
+      repositories(
+        first: 100
+        orderBy: { field: NAME, direction: ASC }
+        after: $endCursor
+      ) {
+        totalCount
+        pageInfo {hasNextPage endCursor}
+        nodes {
+          nameWithOwner
+          repoWide: object(expression: "HEAD:.github/copilot-instructions.md") {
+            ... on Blob {
+              byteSize
+            }
+          }
+          pathSpecific: object(expression: "HEAD:.github/instructions/") {
+            ... on Tree {
+              entries {
+                path
+              }
+            }
+          }
+        }
+      }
+    }
+  }
+' --jq '.data.organization.repositories.nodes[] | select(.repoWide == null and .pathSpecific == null) | .nameWithOwner'