diff --git a/support/OWNERS b/support/OWNERS
new file mode 100644
index 0000000..6890b59
--- /dev/null
+++ b/support/OWNERS
@@ -0,0 +1,12 @@
+# See the OWNERS docs: https://git.k8s.io/community/contributors/guide/owners.md
+
+approvers:
+  - harche
+  - mrunalp
+  - wking
+reviewers:
+  - harche
+  - mrunalp
+  - wking
+
+component: "Support Lightspeed Skills"
diff --git a/support/README.md b/support/README.md
new file mode 100644
index 0000000..1c63515
--- /dev/null
+++ b/support/README.md
@@ -0,0 +1,5 @@
+# Support skills
+
+This directory contains skills which are designed to help agents interact with Red Hat support systems such as Jira.
+
+> **Note:** The skills in this directory are initial drafts. They will evolve as we test and refine them based on real-world usage.
diff --git a/support/jira/SKILL.md b/support/jira/SKILL.md
new file mode 100644
index 0000000..31b6e5d
--- /dev/null
+++ b/support/jira/SKILL.md
@@ -0,0 +1,61 @@
+---
+name: jira
+description: Interact with Red Hat Jira (issues.redhat.com) via REST API. Use when the user asks about Jira issues, shares a Jira URL, mentions issue keys like OCPBUGS-*, OCPNODE-*, or any issues.redhat.com link.
+allowed-tools: Bash(curl:*)
+---
+
+# Red Hat Jira
+
+Interact with Red Hat Jira via REST API with `curl`. Uses a Bearer token from the `JIRA_API_TOKEN` environment variable.
+
+## Authentication
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" "<url>" | python3 -m json.tool
+```
+
+**PAT limitation**: `currentUser()` in JQL and `/rest/api/2/myself` do NOT work with PATs on Red Hat Jira. To find the token owner's username:
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/auth/1/session" | python3 -m json.tool
+```
+
+Then use the `name` field (e.g., `harpatil@redhat.com`) in JQL queries like `assignee="harpatil@redhat.com"`.
+
+## Quick Start
+
+```bash
+# View an issue
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/OCPNODE-4151?fields=summary,status,assignee,priority,issuetype,description,created,updated,components,labels,fixVersions" \
+  | python3 -m json.tool
+
+# Search with JQL
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/search?jql=assignee%3D%22harpatil%40redhat.com%22+AND+type%3DEpic+AND+status+not+in+(Closed,Done)&maxResults=10&fields=summary,status,assignee,priority" \
+  | python3 -m json.tool
+
+# Get children of an epic
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/search?jql=%22Epic+Link%22%3DOCPNODE-4151&fields=summary,status,assignee,issuetype,priority" \
+  | python3 -m json.tool
+```
+
+## URL Parsing
+
+**Jira URLs** like `https://issues.redhat.com/browse/OCPNODE-4151` — extract the issue key (`OCPNODE-4151`) from the path after `/browse/`.
+
+## References
+
+Detailed command references — read on demand:
+
+|references/jira.md — Issues, JQL search, epics, comments, transitions, custom fields
+
+## Important
+
+- **Always include clickable Jira URLs** when displaying issues, epics, or any Jira items. Format: `https://issues.redhat.com/browse/{KEY}`.
+- **Always confirm with the user before creating/updating Jira issues, adding comments, transitioning status, or any write operation.**
+- Use `fields=` parameter to limit Jira response size — full issue responses are very large.
+- Epic children are found via JQL `"Epic Link"=EPIC-KEY`, NOT `parentEpic`.
+- All actions happen as the token owner.
diff --git a/support/jira/references/jira.md b/support/jira/references/jira.md
new file mode 100644
index 0000000..3518ed0
--- /dev/null
+++ b/support/jira/references/jira.md
@@ -0,0 +1,308 @@
+# Jira (issues.redhat.com)
+
+View, search, create, update, and manage Jira issues via the REST API v2.
+
+> **Auth:** Always set the token from the OS secret store before any curl call:
+
+## Base URL
+
+```
+https://issues.redhat.com/rest/api/2
+```
+
+## View Issue
+
+```bash
+# Basic issue details
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}?fields=summary,status,assignee,priority,issuetype,description,created,updated,components,labels,fixVersions,versions,reporter,resolution,resolutiondate,duedate" \
+  | python3 -m json.tool
+
+# With field name mappings (discover custom field names)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}?expand=names" \
+  | python3 -m json.tool
+
+# With changelog (issue history)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}?expand=changelog" \
+  | python3 -m json.tool
+
+# Rendered fields (description/comments as HTML)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}?expand=renderedFields&fields=description,comment" \
+  | python3 -m json.tool
+```
+
+## Epic Children
+
+**Use `"Epic Link"` in JQL** — the `parentEpic` function does NOT work on this instance.
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/search?jql=%22Epic+Link%22%3D{epicKey}&fields=summary,status,assignee,issuetype,priority&maxResults=50" \
+  | python3 -m json.tool
+```
+
+## Comments
+
+```bash
+# List comments
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}/comment" \
+  | python3 -m json.tool
+
+# List comments with pagination
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}/comment?startAt=0&maxResults=10" \
+  | python3 -m json.tool
+
+# Add a comment (ALWAYS confirm with user first)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X POST "https://issues.redhat.com/rest/api/2/issue/{issueKey}/comment" \
+  -d '{"body": "Comment text here"}' \
+  | python3 -m json.tool
+```
+
+## Transitions (Workflow)
+
+```bash
+# List available transitions
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}/transitions" \
+  | python3 -m json.tool
+
+# Perform a transition (ALWAYS confirm with user first)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X POST "https://issues.redhat.com/rest/api/2/issue/{issueKey}/transitions" \
+  -d '{"transition": {"id": "TRANSITION_ID"}}' \
+  | python3 -m json.tool
+```
+
+Note: Some issues (especially Bugzilla-synced bugs) may return empty transitions if the workflow is managed externally.
+
+## Update Issue
+
+```bash
+# Update fields (ALWAYS confirm with user first)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X PUT "https://issues.redhat.com/rest/api/2/issue/{issueKey}" \
+  -d '{
+    "fields": {
+      "summary": "Updated summary",
+      "description": "Updated description",
+      "assignee": {"name": "username@redhat.com"},
+      "priority": {"name": "Major"},
+      "labels": ["label1", "label2"]
+    }
+  }'
+
+# Update a single field
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X PUT "https://issues.redhat.com/rest/api/2/issue/{issueKey}" \
+  -d '{"fields": {"assignee": {"name": "username@redhat.com"}}}'
+```
+
+## Create Issue
+
+```bash
+# Create a new issue (ALWAYS confirm with user first)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X POST "https://issues.redhat.com/rest/api/2/issue" \
+  -d '{
+    "fields": {
+      "project": {"key": "OCPNODE"},
+      "issuetype": {"name": "Story"},
+      "summary": "Issue summary",
+      "description": "Detailed description",
+      "assignee": {"name": "username@redhat.com"},
+      "priority": {"name": "Major"},
+      "customfield_12311140": "EPIC-KEY"
+    }
+  }' | python3 -m json.tool
+```
+
+Common issue types: `Epic`, `Story`, `Task`, `Bug`, `Sub-task`, `Feature`, `Spike`
+
+## Link Issues
+
+```bash
+# Create a link between issues (ALWAYS confirm with user first)
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X POST "https://issues.redhat.com/rest/api/2/issueLink" \
+  -d '{
+    "type": {"name": "Blocks"},
+    "inwardIssue": {"key": "OCPNODE-100"},
+    "outwardIssue": {"key": "OCPNODE-200"}
+  }'
+```
+
+Common link types: `Blocks`, `Cloners`, `Duplicate`, `Relates`
+
+## Watchers
+
+```bash
+# Get watchers
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}/watchers" \
+  | python3 -m json.tool
+
+# Add yourself as a watcher
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -X POST "https://issues.redhat.com/rest/api/2/issue/{issueKey}/watchers" \
+  -d '"username@redhat.com"'
+```
+
+## User Search
+
+Resolve a display name to a Jira username (for use in `assignee =` JQL clauses):
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/user/search?username=First+Last" \
+  | python3 -c "
+import sys, json
+for u in json.load(sys.stdin):
+    print(f'{u[\"displayName\"]:<30} {u[\"name\"]}')"
+```
+
+The `name` field is the Jira username. Use it in JQL: `assignee = "username"`.
+
+## JQL Search
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/search?jql=<URL-ENCODED-JQL>&fields=summary,status,assignee,priority,issuetype&maxResults=20" \
+  | python3 -m json.tool
+```
+
+**Tip:** URL-encode JQL queries. Use `+` for spaces, `%3D` for `=`, `%27` for `'`, `%22` for `"`.
+
+### JQL Operators
+
+| Operator | Example |
+|---|---|
+| `=` | `status = "In Progress"` |
+| `!=` | `status != Done` |
+| `IN` | `status IN ("To Do", "In Progress")` |
+| `NOT IN` | `priority NOT IN (Blocker, Critical)` |
+| `IS EMPTY` | `assignee IS EMPTY` |
+| `IS NOT EMPTY` | `fixVersion IS NOT EMPTY` |
+| `~` (contains) | `summary ~ "kubernetes"` |
+| `>`, `<`, `>=`, `<=` | `created >= "2026-01-01"` |
+| `WAS` | `status WAS "In Progress"` |
+| `CHANGED` | `status CHANGED FROM "To Do" TO "In Progress"` |
+
+### Common Queries
+
+```bash
+# My open epics
+jql: assignee = "harpatil@redhat.com" AND type = Epic AND resolution = Unresolved ORDER BY priority DESC
+
+# Children of an epic
+jql: "Epic Link" = OCPNODE-4151
+
+# Critical bugs updated recently
+jql: project = OCPBUGS AND type = Bug AND priority IN (Blocker, Critical) AND updated >= -7d ORDER BY priority DESC
+
+# Text search
+jql: project = OCPNODE AND text ~ "TLS profile"
+
+# By component
+jql: project = OCPBUGS AND component = "Management Console"
+
+# By fix version
+jql: project = OCPNODE AND fixVersion = "4.22"
+
+# Created this week
+jql: project = OCPNODE AND created >= startOfWeek()
+
+# Unresolved in current sprint
+jql: sprint in openSprints() AND project = OCPNODE AND resolution = Unresolved
+```
+
+### Pagination
+
+```bash
+# First page
+"https://issues.redhat.com/rest/api/2/search?jql=...&startAt=0&maxResults=20&fields=summary,status"
+
+# Next page
+"https://issues.redhat.com/rest/api/2/search?jql=...&startAt=20&maxResults=20&fields=summary,status"
+```
+
+### Tabular Output
+
+```bash
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/search?jql=project%3DOCPNODE+AND+type%3DEpic&maxResults=10&fields=summary,status,assignee,priority" \
+  | python3 -c "
+import sys, json
+data = json.load(sys.stdin)
+print(f'Total: {data[\"total\"]} issues')
+print(f'{\"Key\":<20} {\"Status\":<15} {\"Assignee\":<25} {\"Summary\"}')
+print('-' * 90)
+for issue in data['issues']:
+    f = issue['fields']
+    assignee = f.get('assignee', {})
+    print(f'{issue[\"key\"]:<20} {f[\"status\"][\"name\"]:<15} {(assignee or {}).get(\"displayName\", \"Unassigned\"):<25} {f[\"summary\"][:50]}')
+"
+```
+
+## Field Selection
+
+Use `fields=` to limit response size:
+
+```bash
+# Minimal
+fields=summary
+
+# Standard view
+fields=summary,status,assignee,priority,issuetype,created,updated
+
+# With custom fields
+fields=summary,status,assignee,customfield_12311140,customfield_12310243
+```
+
+## Useful Custom Fields
+
+Red Hat-specific custom fields on issues.redhat.com:
+
+| Custom Field ID | Name | Notes |
+|---|---|---|
+| `customfield_12311140` | Epic Link | Set on children to link to parent epic |
+| `customfield_12311141` | Epic Name | Short name for the epic |
+| `customfield_12311142` | Epic Status | Epic-specific status |
+| `customfield_12310940` | Sprint | Sprint association |
+| `customfield_12310243` | Story Points | Estimation points |
+| `customfield_12313140` | Parent Link | Parent issue link |
+| `customfield_12313240` | Team | Team assignment |
+| `customfield_12313941` | Target start | Planned start date |
+| `customfield_12313942` | Target end | Planned end date |
+| `customfield_12315948` | QA Contact | QA assignee |
+| `customfield_12316142` | Severity | Bug severity |
+| `customfield_12318341` | Feature Link | Link to parent feature |
+| `customfield_12319940` | Target Version | Target release version |
+| `customfield_12320845` | Color Status | RAG status |
+| `customfield_12316840` | Bugzilla Bug | Linked Bugzilla ID |
+| `customfield_12317313` | Release Note Text | Release note content |
+| `customfield_12316749` | Architect | Architect contact |
+| `customfield_12316752` | Product Manager | PM contact |
+| `customfield_12310220` | Git Pull Request | Linked PRs |
+| `customfield_12316542` | Ready | Ready for development |
+| `customfield_12316543` | Blocked | Blocked flag |
+| `customfield_12316544` | Blocked Reason | Reason for block |
+
+```bash
+# Discover all field names for an issue
+curl -s -H "Authorization: Bearer $JIRA_API_TOKEN" \
+  "https://issues.redhat.com/rest/api/2/issue/{issueKey}?expand=names" \
+  | python3 -c "import sys,json; data=json.load(sys.stdin); [print(f'{k}: {v}') for k,v in sorted(data.get('names',{}).items())]"
+```