Install
$ agentstack add skill-danmcinerney-orchflows-orch-frontier ✓ 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
Require: a run's ticket directory forming a finite acyclic dependency graph, and the run's bound — from the spec, or named by the caller for an ad-hoc set.
Open by dispatching the whole ready frontier — every ticket whose depends_on are all complete — in parallel through orch-task, one dispatch per ticket, no two sharing a write scope; a ticket waiting on a dependency stays pending. Then recompute on every event — a result landing, a suspension parking its item, a claim's lease expiring — never on a schedule of rounds: record what landed in the worklog — the tickets alone are the record when the run keeps none; reclaim stale claims per [rules/delegation.md](../../../rules/delegation.md) §11 (a parked item's claim never goes stale); promote each pending ticket whose depends_on are now all complete to ready; set each pending ticket depending on a failed, blocked, or limited ticket to blocked, naming its blocker — a failure blocks exactly its dependents, the rest of the graph rolls on; dispatch everything newly ready immediately. A suspension parks its item at the event step — neither complete nor failed, its dependents wait: the caller satisfies the excluded action at that step and re-readies the ticket; a caller that cannot satisfy it exits with the parked remainder — resume is its own caller's re-dispatch, never this engine's. The engine exits when no ticket is ready or pending and no live dispatch remains — parked items return in the open remainder — and exits limited when the run bound is spent with tickets still open; bounds inherit downward.
Never: start a dependent before its dependencies are complete; hold a ready ticket back to batch it with others; hide a blocked subtree in a summary of the successes; re-order the graph to dodge a failure.
Return: per-ticket results, the graph's terminal state, and the open remainder with what blocks it.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: DanMcInerney
- Source: DanMcInerney/orchflows
- 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.