trace
Bind the current session to a Trace task, and re-enter prior task context. Use when the user names a specific piece of work they are starting, resuming, or continuing — including "scope out X", "define a new X", "plan the work for X", "build X", "I need to build X from scratch", "tackle X", "add X to Y", "work on X", or "get back to X". Covers features, bugs, and refactors. Trace fires FIRST whenever the user commits to a new, named piece of work — whether or not they say "feature" or "task", and even when they also want to brainstorm, scope, or plan it — so the session is bound before any planning or creative skill runs. Does NOT apply when actively debugging or investigating a running system ("this endpoint is throwing errors", "help me debug X"), nor when the user is explicitly deferring the build to explore options first ("before we build anything, let's explore the design space") — neither is a committed new piece of work. Also use when explicitly asked to bind a session to a task, or when session-start context reports no active task during real project work.
Works with
---
name: trace
description: Bind the current session to a Trace task, and re-enter prior task context. Use when the user names a specific piece of work they are starting, resuming, or continuing — including "scope out X", "define a new X", "plan the work for X", "build X", "I need to build X from scratch", "tackle X", "add X to Y", "work on X", or "get back to X". Covers features, bugs, and refactors. Trace fires FIRST whenever the user commits to a new, named piece of work — whether or not they say "feature" or "task", and even when they also want to brainstorm, scope, or plan it — so the session is bound before any planning or creative skill runs. Does NOT apply when actively debugging or investigating a running system ("this endpoint is throwing errors", "help me debug X"), nor when the user is explicitly deferring the build to explore options first ("before we build anything, let's explore the design space") — neither is a committed new piece of work. Also use when explicitly asked to bind a session to a task, or when session-start context reports no active task during real project work.
license: MIT
---
# Trace
Use this skill when the user names a specific feature, fix, or task they are
starting or about to do — even when phrased as "scope out X", "define a new X",
"plan the work for X", "build X from scratch", "tackle X", or "add X to Y".
Trace binds the session before any planning skill runs; other skills do the
subsequent planning or execution. Also use when the user asks to bind the
current session to a task, or when session-start context reports there is no
active task for this session.
## First, follow your host's binding flow
Trace runs in Claude Code, Codex, Cursor, and GitHub Copilot CLI. The hosts differ only in how a
session is identified and in how the "no active task" signal arrives — the
verbs below are the same in all of them. Before binding, read the file for your
host and follow it:
- **Claude Code** — you run with `CLAUDE_CODE_SESSION_ID` in the environment and
receive a SessionStart line about this session's task state. Read
[resources/claude.md](resources/claude.md).
- **Codex** — you run with `CODEX_THREAD_ID` in the environment and have no
session-start hook. Read [resources/codex.md](resources/codex.md).
- **Cursor** — no session env var and no session-start hook; the CLI resolves
your session from the directory it runs in. Read
[resources/cursor.md](resources/cursor.md).
- **GitHub Copilot CLI** — lifecycle hooks register the session and check state
freshness; the CLI infers its identity from Copilot's live-session lock. Read
[resources/copilot.md](resources/copilot.md).
## We're working on X
Resolve an existing task by slug or title `X`, or create one when absent, then
bind the current session to it. New task titles should be human-readable
sentence case ("Break stop and stale expiry"), not kebab-case — the slug is
derived automatically.
Starting to **scope or define** a piece of work counts as working on it: when
the user begins scoping, speccing, or defining a feature (e.g. "let's define
X", "let's scope X", or invoking a planning skill on X), run this same verb
with the feature as the title. The printed `taskDocsDir:` is then in
conversation for downstream skills that produce planning artifacts, so they
know where to write without any extra wiring.
When this **creates** a new task, write a one-line `--description` from the
conversation context — what the work actually is, in the user's terms — so the
task is recognisable later (this is the text the recall skill and the task page
surface). Draft it yourself from what you know; don't stop to interview the
user for it.
```sh
trace skill work-on-task "X" --description "Rework the checkout into a multi-step wizard"
```
`--description` only seeds a freshly created task; when the task already
exists, drop the flag (tending an existing description on re-entry is handled by
the `trace-reenter` skill).
By default the task keys to the project root resolved from the CLI's working
directory. When the work clearly lives in a **different** project than where the
CLI is running — e.g. you are in a multi-project sandbox or wrapper directory but
the task belongs to a specific repo — pass `--project <dir>` pointing at that
project so the task keys to its git root instead of cwd's:
```sh
trace skill work-on-task "X" --project /path/to/that/repo
```
Default to cwd (omit the flag) unless you have a concrete reason the work belongs
to another project. A nonexistent `--project` path is a hard error. The same
`--project <dir>` flag is accepted by `trace task create` and `trace task
capture` for the same reason.
The command prints `taskDocsDir: <path>`. Put task-specific decision docs,
plans, handoffs, and notes in that directory so future re-entry sees them
without a separate registration step.
## Re-enter a task
When the user names an exact task slug or title to resume, hand off to the
`trace-reenter` skill, which owns the re-entry flow — it fetches the manifest
and binds the current session atomically. Don't issue a separate `work-on-task`
bind after re-entry.More Observability skills
google-agents-cli-observability
google/agents-cli
>
azure-observability
microsoft/azure-skills
Azure Observability Services including Azure Monitor, Application Insights, Log Analytics, Alerts, and Workbooks. Provides metrics, APM, distributed tracing, KQL queries, and interactive reports. USE FOR: Azure Monitor, Application Insights, Log Analytics, Alerts, Workbooks, metrics, APM, distributed tracing, KQL queries, interactive reports, observability, monitoring dashboards. DO NOT USE FOR: instrumenting apps with App Insights SDK (use appinsights-instrumentation), querying Kusto/ADX clusters (use azure-kusto), cost analysis (use azure-cost-optimization).
social
coreyhaines31/marketingskills
When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, Facebook, or other platforms, or wants to do social listening and engagement triage. Also use when the user mentions 'LinkedIn post,' 'Twitter thread,' 'social media,' 'content calendar,' 'social scheduling,' 'engagement,' 'viral content,' 'what should I post,' 'repurpose this content,' 'tweet ideas,' 'LinkedIn carousel,' 'social media strategy,' 'grow my following,' 'TikTok video,' 'Reels,' 'Shorts,' 'video script,' 'video hook,' 'short-form video,' 'create a reel,' 'social listening,' 'brand mentions,' 'competitor monitoring,' 'top posts to comment on,' 'find people asking for,' 'carousel,' 'slide-by-slide,' or 'document post.' Use this for social media content creation, repurposing, scheduling, short-form video scripting, and social listening. For broader content strategy, see content-strategy. For paid ads, see ad-creative. For earned media, see public-relations.

