Skip to main content
Glama

mcp-posthog__agent-feedback

Read-only

Send feedback about anything PostHog to the PostHog team. Set feedback_type to route it: product (any PostHog product or feature — insights, session replay, feature flags, the data warehouse, web analytics, error tracking, etc.; put the area in product_area), mcp (this MCP server itself — a tool, input schema, response format, error, or these instructions; set category), docs, scout (reserved for scheduled scout runs reporting an improvement opportunity in the canonical PostHog scout skill steering the run; set scout_skill_name, scout_skill_version, and scout_category, and generalize — describe the pattern, never the project's data), or other. ALL sentiments are welcome via sentiment (positive/neutral/negative/mixed) — praise and feature requests are useful signal, not just problems. Good triggers: a confusing or broken product experience, a papercut that slowed the task down, a missing capability you worked around, a feature request, an unhelpful error, or something that work… Through Royally: its server marks it read-only, so it runs straight away.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "summary": "<summary>",
      "feedback_type": "<feedback_type>",
      "sentiment": "<sentiment>",
      "context": "<context>"
    },
    "name": "mcp-posthog__agent-feedback"
  }
}

Input Schema

NameRequiredDescriptionDefault
contextYesA few words on why you are calling this tool, for the server developer's analytics. Describe people and things by role, such as "a customer". Do not quote the user or include names, emails, or other personal details.
detailsNoAny additional context that doesn't fit the other fields. Keep it to clear, concise bullet points.
site_idNoSite to run this tool on: 5c727948-86e8-4a18-a2fe-2b6b99c6837a (https://mcp.posthog.com/mcp?mode=tools&features=web_analytics,error_tracking)
summaryYesA one-sentence headline capturing the feedback (e.g. "session replay scrubber jumps backwards when you click the timeline", "query-trends descriptions made it hard to choose between trends and funnels", or "the new SQL editor autocomplete is excellent").
categoryNoFor MCP feedback (`feedback_type: "mcp"`) only: the single category that best describes the dominant theme. Pick "missing_tool" if a capability was absent, "tool_description" if the tool docs were unclear, "tool_input_schema" if input args were confusing, "tool_output_format" if the response was hard to consume, "instructions_clarity" if these MCP instructions were unclear, "tool_correctness" if a tool returned wrong data, "error_message" if an error was unhelpful, "performance" if latency was the issue. Omit for product, docs, or other feedback.
llm_modelNoThe model identifier you are running as, for example "claude-opus-4-8" or "gpt-5.2". Pass "unknown" if you are not certain. Recorded for the server developer's analytics.
sentimentYesThe overall tone. Use "negative" for something broken or blocking, "mixed" for mostly-fine-but-with-a-concrete-problem, "neutral" for a suggestion or feature request with no strong sentiment, and "positive" for praise or something that worked well. All sentiments are welcome — positive feedback is encouraged, not just problems.
tools_usedNoThe MCP tool names you called while working on the user's task (e.g. ["read-data-schema", "query-trends"]). Helps us correlate feedback to specific tools.
product_areaNoThe PostHog product or area this is about, in free text (e.g. "session replay", "insights", "data warehouse", "feature flags", "docs"). Most useful for product feedback; for MCP feedback the tool name belongs in `details`/`friction_points` instead.
user_requestNoA short, anonymised paraphrase of what the user originally asked you to do. Do not include PII, customer names, or sensitive query content.
feedback_typeYesWhat this feedback is about. "product" = any PostHog product or feature (insights, session replay, feature flags, the data warehouse, web analytics, error tracking, etc.). "mcp" = this MCP server itself — a tool, its input schema, response format, an error, or these instructions. "docs" = PostHog documentation. "scout" = a canonical PostHog scout skill's content — reserved for scheduled scout runs reporting an improvement opportunity in their own PostHog-authored skill (set `scout_skill_name`, `scout_skill_version`, and `scout_category`). "other" = anything that doesn't fit the above.
scout_categoryNoFor scout feedback (`feedback_type: "scout"`) only: the single category that best describes the skill gap. "false_positive" = the skill's detection rules surfaced something that wasn't real; "missed_detection" = a real issue the skill's instructions steered you past; "discriminator_gap" = the skill's signal-vs-baseline discriminator doesn't hold for a class of projects; "wasted_investigation" = an investigation pattern the skill mandates burned budget without payoff; "instruction_ambiguity" = an instruction that is ambiguous in practice. Omit for non-scout feedback.
task_completedNoWere you able to complete the user's task? Be honest — "false" is just as useful as "true". Most relevant when `feedback_type` is "mcp".
conversation_idNoThe conversation_id this server returned earlier in this conversation, unchanged, including after later user messages. Omit it if you have none. Recorded for the server developer's analytics.
friction_pointsNoClear, concise bullet points describing the friction — what was confusing, broken, slow, or missing. Quote the exact product surface, tool name, parameter, or error text where you can. Omit for purely positive feedback.
scout_skill_nameNoFor scout feedback (`feedback_type: "scout"`) only: the canonical scout skill the feedback is about (e.g. "signals-scout-web-analytics"), exactly as named in the run identity. Required for scout feedback — without it the feedback cannot be aggregated per skill.
scout_skill_versionNoFor scout feedback (`feedback_type: "scout"`) only: the skill version the run executed (from the run identity). Feedback is only actionable against the version that produced it — the skill may have moved since.
suggested_improvementNoThe single most impactful, concrete change that would address this feedback, if you can name one (e.g. "add a `filters` example to query-funnel's description", or "let the replay scrubber snap to the nearest event"). Optional — praise or an observation doesn't need one.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A3.8/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=true, destructiveHint=false and openWorldHint=false, so the safety profile is covered and the description's restatement ("its server marks it read-only, so it runs straight away") adds nothing. Beyond annotations the description explains routing destinations and some privacy expectations, but never says what happens after submission — confirmation, whether feedback is tied to an account, or any rate limit.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The opening is well front-loaded and the per-type routing list is efficient. However the sentence that begins "Good triggers: …" is cut off mid-word and is followed by an incoherent fragment ("Through Royally: its server marks it read-only…"), which is noise an agent must ignore.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For an 18-parameter submission tool with a fully documented schema and no output schema, the definition covers the decision-relevant parts: what each feedback_type means, which fields pair with which type, and what counts as valid feedback. Only the post-submission behavior is left unspecified.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and all 18 parameters carry their own descriptions, including conditional guidance ("For MCP feedback only", "Omit for non-scout feedback"). The description's condensed restatement of feedback_type values and the dependent fields (product_area, category, scout_skill_name/version/category) mirrors the schema rather than adding new meaning, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource up front — "Send feedback about anything PostHog to the PostHog team" — and immediately enumerates the routing dimension (product/mcp/docs/scout/other), so an agent knows exactly what this tool is for versus every other PostHog tool.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit triggers (confusing or broken experience, papercut, missing capability, unhelpful error, feature request) and explicitly welcomes positive sentiment, which is real when-to-use guidance. It stops short of saying when NOT to use it (e.g. account/support issues) or naming an alternative channel, so it is clear context but not full exclusion coverage.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources