mirror of
https://github.com/getsentry/sentry-for-ai.git
synced 2026-09-14 14:28:22 +08:00
Evan Purkhiser
cc302403eb
Task-shaped skills will draw their reference material (per-platform SDK guides, per-signal concepts) from a single maintained library rather than duplicating it. Each shipped skill must still be self-contained, so the build copies a skill's declared references into it. Add a CONTENT_ROOT env var (default: repo root) so the same build steps can assemble an alternate skill tree, e.g. CONTENT_ROOT=skills-next. hydrate-references.py walks a skills directory: for every skill with a references.yml it brace-expands and globs the `needs:` patterns against the library, copies the matches into the skill's references/ preserving subpaths, and strips the manifest so it never ships. It runs via `uv run --script` with inline PyYAML and braceexpand dependencies. build-common.sh invokes it once per build, and only when some skill declares references, so existing skills are unaffected and the common case spawns no interpreter. Factor the steps shared across all four agent builders into build-common.sh (content-root resolution, skill copy+hydrate, commands, SKILL_TREE), sourced by each plugin-src/<agent>/build.sh.
42 lines
1.7 KiB
Bash
Executable File
42 lines
1.7 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
#
|
|
# build.sh — Build the Claude Code distribution of the Sentry plugin.
|
|
#
|
|
# Claude consumes the plugin at the dist branch ROOT (its
|
|
# `.claude-plugin/marketplace.json` uses `source: "./"`) and HONORS
|
|
# `disable-model-invocation`, so the skill-tree routing works natively. No skill
|
|
# mutation is needed: this copies the manifest (from alongside this script) plus
|
|
# the shared content from the repo root.
|
|
#
|
|
# Claude's MCP server is declared inline in `plugin.json` (`mcpServers`), so no
|
|
# `.mcp.json` is shipped here; the root `mcp.json` only feeds the Cursor and
|
|
# Codex builds.
|
|
#
|
|
# Skill content (skills/, commands/, references/, SKILL_TREE.md) is read from
|
|
# CONTENT_ROOT, defaulting to the repo root. Point it at an alternate tree (e.g.
|
|
# CONTENT_ROOT=skills-next) to build a different skill set with the same steps.
|
|
#
|
|
# Usage: build.sh <TARGET_DIR> (TARGET_DIR assumed empty)
|
|
|
|
set -euo pipefail
|
|
|
|
TARGET_DIR="${1:?usage: build.sh <TARGET_DIR>}"
|
|
SRC_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
REPO_ROOT="$(cd "$SRC_DIR/../.." && pwd)"
|
|
cd "$REPO_ROOT"
|
|
source "$REPO_ROOT/scripts/build-common.sh"
|
|
resolve_content_root "$REPO_ROOT"
|
|
|
|
mkdir -p "$TARGET_DIR/.claude-plugin"
|
|
|
|
cp "$SRC_DIR/plugin.json" "$TARGET_DIR/.claude-plugin/plugin.json"
|
|
cp "$SRC_DIR/marketplace.json" "$TARGET_DIR/.claude-plugin/marketplace.json"
|
|
copy_skills "$CONTENT_ROOT" "$TARGET_DIR/skills"
|
|
copy_commands "$CONTENT_ROOT" "$TARGET_DIR/commands"
|
|
copy_skill_tree "$CONTENT_ROOT" "$TARGET_DIR/SKILL_TREE.md"
|
|
rsync -a assets/ "$TARGET_DIR/assets/"
|
|
cp "$SRC_DIR/README.md" "$TARGET_DIR/README.md"
|
|
cp LICENSE "$TARGET_DIR/LICENSE"
|
|
|
|
echo "Built Claude dist into $TARGET_DIR (root plugin, content from $CONTENT_ROOT)."
|