api-patterns
3CX's native PBX MCP server: the per-PBX endpoint shape (every PBX is its own FQDN and its own OAuth authorization server — there is no shared mcp.3cx.com),…
Read-only lookups of PostHog early-access feature flags and experiments — rollout status, targeting, and configuration. Does not create, update, or delete flags or experiments.
$ npx -y skills add wyre-technology/msp-claude-plugins --skill feature-flags-and-experiments --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/feature-flags-and-experimentsContext preview
The summary Claude sees to decide when to auto-load this skill.
Read-only lookups of PostHog early-access feature flags and experiments — rollout status, targeting, and configuration. Does not create, update, or delete flags or experiments.
name: "PostHog Feature Flags & Experiments" description: > Read-only lookups of PostHog early-access feature flags and experiments — rollout status, targeting, and configuration. Does not create, update, or delete flags or experiments. when_to_use: >- When checking whether a feature flag or early-access feature is live for a client, or reviewing an experiment's configuration. Use when: posthog feature flag, posthog experiment, early access feature, rollout percentage, flag status, feature flag targeting, or a/b test status.
Feature flags gate whether a piece of code runs for a given user or percentage of traffic. Early-access features are PostHog's primitive for managing a beta opt-in program — customers who've asked to try something before general availability. Experiments run a controlled comparison between variants of a flag against a primary metric. For an MSP, the recurring question this skill answers is "is this feature actually live for this client" or "what is this experiment currently testing" — not turning either on or off.
or experiment** — this skill and this plugin are read-only. See [GOVERNANCE.md](../../GOVERNANCE.md), *Tool permission tiers*: feature-flag mutations change what code path a live client's production application executes, immediately and with no built-in dry run, which is exactly why this plugin excludes them at v1. Make the change directly in the PostHog UI, with the same care you'd give any production config change.
`cohorts-and-events`; this skill only reports what a flag or experiment is configured to target.
A feature flag has a key, a rollout condition (percentage, specific users, or property-based targeting), and an enabled/disabled state per condition group. Early-access features are a distinct PostHog primitive layered on top of flags, purpose-built for beta programs — a user opts in, and that opt-in maps to a flag being active for them. Experiments attach two or more flag variants to a primary metric and report which variant is winning.
The confirmed read tool family for this domain:
Both are read-only; the corresponding write tools (`early-access-feature-create`, `early-access-feature-destroy`, `early-access-feature-partial-update`) and general feature-flag CRUD are excluded from this plugin — see [GOVERNANCE.md](../../GOVERNANCE.md). Experiment lookups (list/get) follow the same read-only pattern. PostHog's own docs carry the exhaustive tool catalog, including the broader feature-flag API this plugin does not expose: [posthog.com/docs/model-context-protocol/tools](https://posthog.com/docs/model-context-protocol/tools).
A client asks why a beta feature isn't showing up for one of their users:
1. `early-access-feature-list` to find the feature by name 2. `early-access-feature-retrieve` to check its current status and targeting configuration 3. Report what's configured — if the fix requires changing rollout percentage or targeting, that's a write action outside this plugin's surface; hand it to a human with the appropriate PostHog access
1. List current experiments and their status 2. Pull each experiment's configuration (variants, primary metric) via the read tool 3. Summarize what's actively being tested for the client-facing report — experiment results themselves are analytics data, so cross-reference with `insights-and-dashboards` if the QBR needs the metric outcome, not just the configuration
(killing a broken feature) is outside this plugin. Escalate to a human with direct PostHog access rather than looking for a workaround tool.
tool surface.** The confirmed read tools here are specifically the early-access-feature family. PostHog's general feature-flag API has its own tools, documented separately — don't assume a flag question is automatically answerable through the early-access tools.
them right now** — that's a cohort or event-level question; see `cohorts-and-events`.
One command to supercharge Claude Code for MSP workflows. Then restart Claude Code. That's it. Documentation: mcp.wyre.ai
Repo: wyre-technology/msp-claude-plugins
3CX's native PBX MCP server: the per-PBX endpoint shape (every PBX is its own FQDN and its own OAuth authorization server — there is no shared mcp.3cx.com),…
3CX's live-operations surface: read-only visibility into active calls, recordings, voicemail, department and queue membership, and forwarding/presence…
3CX's read-only directory surface: resolving a caller by email or by exact extension, searching the PBX's own phonebooks, searching contacts synced from an…
3CX's system-and-configuration surface: server time, PBX event log and application log search, service status, database schema and the read-only SELECT-only…
Abnormal Security abuse mailbox cases: user-reported email submissions, case statuses and judgments, the case lifecycle, bulk and remediation actions, and…
Abnormal Security message analysis: message retrieval, email header inspection, attachments, sender reputation, delivery context, and SPF/DKIM/DMARC…