💡 Workflow Execution Protections Adoption Standard

[github-actions[bot]]

Summary

Define an org-wide standard for adopting GitHub's new Workflow Execution Protections (public preview June 18, 2026), which use the rulesets framework to create allow lists controlling who and which events can trigger workflows. This directly addresses the org's existing concerns around unauthorized workflow triggering and agent credential exposure.

Market Signal

GitHub announced Workflow Execution Protections on June 18, 2026 — the latest addition to the 2026 security roadmap. Built on the rulesets framework, this feature introduces two rule types:

• Actor rules: specify who can trigger workflows (individual users, roles like admins, or trusted automation like GitHub Apps, Copilot, Dependabot)
• Event rules: define which GitHub Actions events are permitted (push, pull_request, workflow_dispatch, etc.)

GitHub Actions evaluates these rules before a run starts, so unauthorized actors or events never trigger execution. This represents a shift from distributed per-workflow security configuration to centralized policy management — and is GitHub's direct answer to the "Comment and Control" attack class where untrusted contributors trigger agent workflows via PR events.

Sources: GitHub Changelog (June 18), GitHub 2026 Security Roadmap

User Signal

The org has 30 workflow files and multiple agent-driven workflows (dev-lead, feature-ideation, compliance-audit, agent-shield) that respond to issue, PR, and discussion events. Several existing issues stem directly from uncontrolled workflow triggering:

• Issue dev-lead: issue-pickup runs cancelled by concurrency group → 111 labeled issues stuck with no PR #402: dev-lead concurrency race from label creation events
• Issue fix(dev-lead): multi-label issue creation can cancel the dev-lead dispatch (race) #443: multi-label issue creation cancels dev-lead dispatch
• PR feat: implement issue #465 — Restrict auto-rebase fan-out to review-ready PRs (free, plan-independent mitigation) #468: manual restriction of auto-rebase fan-out to review-ready PRs

These are all manual mitigations for what Workflow Execution Protections solve architecturally.

Technical Opportunity

The org already manages rulesets via apply-rulesets.sh and enforces the pr-quality ruleset across repos. Workflow Execution Protections extend the same rulesets framework, so adoption fits the existing governance model seamlessly. The compliance-audit can be extended to verify execution protection rulesets are configured. The apply-rulesets.sh script can be updated to deploy execution protection rulesets alongside existing ones.

Assessment

Dimension	Score	Rationale
Feasibility	high	Extends existing rulesets infrastructure — apply-rulesets.sh already manages rulesets
Impact	high	Eliminates entire class of workflow triggering bugs and security issues
Urgency	high	Feature just shipped in preview; org has active issues this solves

Adversarial Review

Strongest objection: Also in public preview, and may require Enterprise-tier GitHub plan — not all tiers may have access.
Rebuttal: The org already uses org-level rulesets (pr-quality), confirming the required tier. Execution protections are the natural next layer in the same framework. Even in preview, documenting the intended policy now means the org can adopt immediately when GA arrives. The acute risk of uncontrolled agent workflow triggering (demonstrated by real CVEs) justifies early planning.

Suggested Next Step

Draft a standards/workflow-execution-protections.md defining: required actor rules per workflow category (agent workflows, CI workflows, deploy workflows), required event rules, integration with existing apply-rulesets.sh, and a compliance-audit check for execution protection presence.

[github-actions[bot]]

Weekly Update

What Changed

Workflow Execution Protections entered public preview on June 18, 2026 (GitHub Changelog). This idea was originally proposed when the feature was only on the roadmap — it is now available for immediate adoption.

Key details from the preview:

• Two rule types available: Actor Rules (who can trigger workflows — users, roles, GitHub Apps, Copilot, Dependabot) and Event Rules (which events are permitted — push, pull_request, pull_request_target, workflow_dispatch)
• Built on the rulesets framework: Organization-wide rulesets can target specific repos using repository custom properties
• Evaluate mode available: Rules assess impact without enforcing — surfaces every workflow run that would have been blocked, letting the org assess before activating
• Addresses four attack patterns: Poisoned pipeline execution via PRs, manual-trigger abuse, untrusted-actor execution, and misconfiguration exploitation
• Access path: Settings → Actions → Policies (new section)

Additionally, actions/checkout v7 reached GA with safer pull_request_target defaults that refuse to fetch fork PR code by default. Enforcement will be backported to all supported major versions on July 16, 2026 (GitHub Changelog).

Updated Assessment

Dimension	Previous	Current	Delta
Feasibility	med	high	↑ Feature is now available — evaluate mode allows safe rollout
Impact	high	high	→ Unchanged — still addresses critical CI security gaps
Urgency	med	high	↑ Feature is in preview NOW; evaluate mode available for immediate adoption

Recommendation

Advance. Enable evaluate mode immediately across the org to audit which workflows would be affected. This provides data for an informed enforcement plan with zero risk of disruption. The checkout v7 migration should also begin before the July 16 backport date.

[github-actions[bot]]

Weekly Update

What Changed

Workflow Execution Protections are now in public preview (June 18, 2026). Key developments since this idea was proposed:

• Public preview launched: GitHub shipped "Control who and what triggers GitHub Actions workflows" on June 18, 2026. The feature is built on the ruleset framework the org already uses for pr-quality and code-quality.
• Actor rules: Specify who can trigger workflows — individual users, roles (repository admins), or trusted automation (GitHub Apps, Copilot, Dependabot).
• Event rules: Define which GitHub Actions events are permitted (push, pull_request, workflow_dispatch, etc.).
• Evaluate mode: Rules can run in evaluate mode — every workflow run that would have been blocked is surfaced in policy insights without actually blocking. This eliminates the risk of breaking existing automation.
• Org-wide application: Enterprises can apply consistent policies org-wide using rulesets and repository custom properties.
• Industry validation: GitGuardian and Wiz both published hardening guides recommending immediate adoption. The OWASP ASI framework identifies unauthorized workflow triggering as a supply chain risk vector.

Updated Assessment

Dimension	Previous	Current	Delta
Feasibility	med	high	↑ Feature is in public preview with evaluate mode
Impact	high	high	→ Unchanged — critical for org with 12+ automated workflows
Urgency	med	high	↑ Public preview means adoption can start now

Recommendation

Advance — begin evaluate-mode pilot immediately. The org has 12+ automated workflows triggered by various events. The evaluate mode makes adoption zero-risk. Recommended rollout:

1. Enable workflow execution protections in evaluate mode on petry-projects/.github
2. Monitor policy insights for 2 weeks to discover current trigger patterns
3. Draft actor and event policies for sensitive workflows (canary-rollout, compliance-audit, apply-rulesets)
4. Enforce on .github first, then roll out fleet-wide via apply-rulesets.sh

The org's existing experience with rulesets (pr-quality, code-quality) makes this a natural next step.

[github-actions[bot]]

Weekly Update

What Changed

Workflow Execution Protections reached public preview on June 18, 2026 — this feature is no longer roadmap speculation; it's available for configuration today. Key developments since this discussion was last updated:

1. Two concrete rule types shipped: Actor rules (individual users, repository roles, GitHub Apps, Copilot, Dependabot) and Event rules (push, pull_request, pull_request_target, workflow_dispatch). More rule types are planned.

2. Rulesets framework integration: WEP leverages the existing rulesets infrastructure this org already uses for pr-quality and code-quality. This means the org's apply-rulesets.sh can deploy WEP policies fleet-wide — no new deployment mechanism needed.

3. Shadow/evaluate mode: Rules can be deployed in evaluate mode first, logging what would be blocked without actually blocking it. This maps directly to the org's canary-ring promotion pattern.

4. Configuration at three levels: Enterprise-wide, organization-wide, and repository-specific — using custom properties for repo scoping.

5. Four attack patterns addressed: Poisoned pipeline execution (restrict pull_request_target), manual-trigger abuse (limit workflow_dispatch), untrusted-actor execution (block low-trust identities), and misconfiguration exploitation (centralized policies override workflow files).

Updated Assessment

Dimension	Previous	Current	Delta
Feasibility	med	high	↑ Feature shipped; uses existing rulesets framework
Impact	high	high	→ Unchanged; addresses same threat model
Urgency	med	high	↑ Public preview ready for evaluate-mode deployment today

Recommendation

Advance. This org's agentic workflows (dev-lead, compliance-audit, feature-ideation, canary-rollout) are high-value targets. WEP's evaluate mode enables a zero-risk trial: deploy the ruleset, observe what it would block for one canary cycle, then enforce. Concrete next step: create standards/rulesets/workflow-execution-protections.json template and deploy in evaluate mode to petry-projects/.github first.