Personal Claude Code plugin marketplace — a curated set of reusable skills and agents for full-cycle software development.
Full development cycle toolkit.
| Skill | Invocation | What it does |
|---|---|---|
implement |
/implement [issue or description] |
Full-cycle feature development: planning → atomic commits → validate → review → ship |
validate |
/validate |
Lint, typecheck, targeted tests, and browser check on the current branch |
review |
/review [focus] |
3-angle parallel code review: bugs/security, simplicity, conventions |
ship |
/ship |
Push branch, open PR, wait for CI, clean up worktrees |
plan-issue |
/plan-issue [description] |
Plans a task via Opus agent and creates a GitHub issue with acceptance criteria |
fix-ci |
/fix-ci [PR number] |
Autonomously fixes failing CI: polls checks, reads logs, commits fixes, up to 3 cycles |
Bundled agents: task-planner (Opus, planning), task-reviewer (Sonnet, 3-dimension code review).
Tools for working with Claude itself.
| Skill | Invocation | What it does |
|---|---|---|
skill-review |
/skill-review [skill-name] |
Audits how well a named skill was executed — phase adherence, gate compliance, outcome |
skill-authoring-guide |
auto-loaded | Reference for creating skills, subagents, and hooks (not user-invocable) |
find-docs |
/find-docs |
Fetches current library/framework docs via Context7 CLI |
review-session |
/review-session |
Analyses the current session for improvement opportunities across 7 dimensions — missed skills, delegation decisions, rework, planning gaps, context bloat, parallelisation, and other observations |
Add the marketplace from within Claude Code:
/plugin marketplace add taciogt/ai-plugins
Then install the desired plugins:
/plugin install deliver@taciogt-plugins
/plugin install craft@taciogt-plugins
To get new plugin versions after the repo is updated:
/plugin marketplace update taciogt-plugins
If you're iterating on the plugins locally, add the marketplace by its local path so changes are visible without reinstalling:
/plugin marketplace add /path/to/ai-plugins
Then install plugins the same way:
/plugin install deliver@taciogt-plugins
/plugin install craft@taciogt-plugins
After editing a skill or agent, reload the marketplace to pick up changes:
/plugin marketplace update taciogt-plugins
For contributing guidelines, see CONTRIBUTING.md.
The following /statusline prompt configures a two-line status bar:
- Line 1 — session location: path, git branch, worktree name (when active), model, effort level (when non-default), 💡 (when reasoning enabled)
- Line 2 — performance: color-coded context bar + %, cost (adaptive precision), session time with API portion, rate limits (when available)
/statusline Create a two-line status line:
Line 1 — session location and context:
- Full working directory path shortened with ~ for home, in bold blue
- Git branch in parentheses, in magenta — only when inside a git repo
- Worktree name prefixed with ⑂ — only show when workspace.git_worktree or worktree.name is present
- Model display name, in cyan
- Effort level (e.g. "effort:high") — only show when effort.level is present and not "medium"
- 💡 emoji — only show when thinking.enabled is true
- Segments separated by |
Line 2 — performance metrics:
- 10-block color-coded progress bar (▓ filled, ░ empty) + context percentage + token count in k (e.g. "64k tokens"). Green when under 70%, yellow 70–89%, red at 90%+
- Session cost in USD: 4 decimal places when under $0.01 (e.g. $0.0034), 2 decimal places otherwise (e.g. $1.23)
- Session time: total wall-clock in human-readable format (45s, 12m, 1h32m) with API processing time in parentheses — derived from cost.total_duration_ms and cost.total_api_duration_ms (e.g. "⏱ 12m (3m api)")
- Rate limit usage (e.g. "5h: 23% | 7d: 41%") — only show when rate_limits data is present. Green under 50%, yellow 50–79%, red at 80%+
- Segments separated by |
v0.1.0 — initial release