Install
$ agentstack add skill-intuned-skills-self-healing ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Self-Healing Projects and Issues
Background concept for the fix-open-issues skill. Full docs:
What an Issue is
An Issue is a confirmed problem on a deployed project, they are raised on Job run errors, unexpected results, reduced result size, or an increase on error rate. The platform reaches it through a pipeline:
- Monitoring — after each job run or standalone run, the platform captures metrics: success rate, failure count, run count, and result size, and checks them against anomaly-detection rules and historical data.
- Triage — when an anomaly fires, the agent investigates it (reading healthy and failed runs, traces, and logs) and decides whether it's a real issue or a false positive.
- Surfacing — confirmed issues with the same root cause are grouped and shown in the project's Issues panel.
The issue record
intuned platform issues get --json returns:
- name — short title.
- explanation — a brief, high-level description of what looks wrong and roughly where. It has a description of the issue.
Lifecycle — fixed, dismissed, re-raised
- Issues are OPEN when surfaced.
- You clear an issue by dismissing it — with the CLI (
intuned platform issues dismiss) or the "Dismiss" button in the platform Issues panel. This applies both to false positives and to issues you've fixed. - Dismissing is not fixing — it only removes the issue from the open list. If the root cause is still there, monitoring re-raises the issue on the next job runs.
Platform self-healing levels
Self-healing is configured per-project via intuned platform self-healing set :
| Level | What it does | | --------------- | ----------------------------------------------------------------------------- | | off | Self-healing disabled entirely. | | monitor | Automatically detect and surface issues from runs. | | auto-fix | Automatically open an agent session to fix raised issues on a new branch. | | auto-merge | Automatically merge fix branches. | | auto-deploy | Automatically deploy the project after a fix branch is merged. |
auto-fix, auto-merge, and auto-deploy require a Hosted project. Connected (CLI) projects can only use off or monitor.
In the UI a user can also attach an issue to a message — reference it in the agent input to start a conversation, and the agent loads the issue context and begins investigating.
fix-open-issues is the manual, CLI-side equivalent of auto-fix: it does the discovery → triage → investigate → fix that auto-fix automates, but it stops at a deploy gate so the user owns the deploy (the equivalent of auto-merge / auto-deploy) instead of it happening automatically.
Enabling self-healing (monitor level)
To automatically detect Issues from job runs or standalone runs, set the self-healing level to monitor:
intuned platform self-healing set monitor
From that point on, any run with failing metrics will raise an Issue. Check the current level with intuned platform project get.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Intuned
- Source: Intuned/skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.