Menu
Create annotated milestone or data tags with changelog
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Freeze this session's state for later resumption with claude --resume
List frozen sessions and provide resume commands for all panels
ADMET prediction agent - assess absorption, distribution, metabolism, excretion, and toxicity profiles from molecular structure and physicochemical properties
Drug candidate ranking agent - multi-criteria scoring and prioritization of compounds for Oral Mucositis treatment
Clinical feasibility assessment agent - evaluate practical development pathways, regulatory strategy, cost estimates, and real-world viability for drug candidates
Combination therapy design agent - rational multi-compound strategy design, synergy assessment, and Ayurvedic formulation evaluation
| name | tag-release |
| description | Create annotated milestone or data tags with changelog |
| when_to_use | When creating git tags for version milestones, data collection events, or release tagging |
| allowed-tools | Bash(git tag:*), Bash(git log:*), Bash(git rev-parse:*), Bash(git status:*), Bash(git push:*), Bash(git describe:*), Bash(git diff:*), Bash(git show:*) |
First, reread the following files:
.claude/skills/tag-release/SKILL.md)Create annotated git tags for project milestones and data collection events. This project uses two tag namespaces:
v<major>.<minor>.<patch>
v0.1.0 — Initial data pipeline working (all scrapers functional)
v0.2.0 — Full ChemBL integration (secondary datasets complete)
v0.3.0 — Cross-database integration (ChemBL-IMPPAT mapping at scale)
v0.4.0 — Analysis layer (candidate scoring, ranking)
v0.5.0 — RAG system operational
v1.0.0 — End-to-end pipeline producing research-grade output
data/chembl-full-<YYYY-MM-DD> — After a full ChemBL collection run
data/neo4j-import-<YYYY-MM-DD> — After a successful full graph import
analysis/candidates-v<N> — When candidate list is regenerated
# List existing tags (most recent first)
git tag -l --sort=-creatordate
# If tags exist, show what's changed since the last one
git describe --tags --abbrev=0 2>/dev/null # most recent tag
# Then: git log <last-tag>..HEAD --oneline
Based on the user's input after the command, or by asking:
If the user specifies a version (e.g., v0.1.0):
If the user says "data" or specifies a data tag:
data/chembl-full-2026-04-06If the user doesn't specify:
v0.1.0 as the initial milestone (or a data tag if this is a data collection event)For version tags, create an annotated tag with a changelog:
v0.X.Y — <one-line summary>
Changes since <previous-tag-or-initial>:
- <category>: <description>
- <category>: <description>
...
Data status:
- ChemBL approved drugs: <count> (full/test)
- ChemBL secondary datasets: <status>
- Neo4j graph: <node/relationship counts if known>
- Scrapers migrated: <list>
For data tags, create a shorter annotation:
data/<type>-<date>
Collection details:
- Dataset: <name>
- Records collected: <count>
- Duration: <if known>
- Mode: full/test/incremental
Show the user:
Ask for confirmation before creating.
# Annotated tag (always use -a for this project)
git tag -a <tag-name> -m "<annotation message>"
After creating, show the tag details:
git show <tag-name>
Do NOT automatically push tags to remote. Ask:
"Tag created locally. Push to remote with
git push origin <tag-name>?"
Only push if the user confirms. To push all tags: git push origin --tags
-a) — never lightweight tags. Annotated tags store the tagger, date, and message.Process any arguments passed after the command (tag name, version number, or "data"):