Skip to content

Innovation Tools

AI R&D tools for building agents, skills, and MCP integrations — plus the Slack feedback loop that keeps the work grounded in real user signal. These are the tools we use to build (and improve) the tools.


Current Innovation Skills

inn-seer-agent-skill-developer

Canonical skill for building and improving Seer skills, plugins, hooks, and agent workflows.

Use when you need to:

  • design or refactor Seer skill architecture
  • author or improve plugin structure and activation logic
  • standardize workflow quality across divisions

Skill Reference


inn-slack-research

Structured Slack research with two distinct modes that share a common methodology.

Mode 1 — Feedback Loop: on-demand sweep that captures organic feedback on Seer's AI skills, agents, and tools, classifies signals (bug / request / positive / gap / adoption), and produces a reviewable digest. Includes urgency calibration, narrative framing, and Wrike routing guidance.

Mode 2 — Prototype Discovery: targeted search that finds and ranks prototypes, experiments, and built tools shared in Slack — for migration into agents-infra or documentation.

Use when you need to:

  • run an on-demand feedback sweep for a specific skill or agent
  • identify historical implementation patterns before rebuilding
  • find high-signal internal discussions and examples
  • gather evidence for innovation prioritization and roadmap direction

Resources:

  • resources/feedback-sweep.md — full execution guide (channel map, queries, taxonomy, urgency rubric, narrative framing, digest template)
  • resources/wrike-routing.md — signal → Wrike action map with folder IDs, item type IDs, and description templates
  • resources/prototype-discovery.md — 115-point scoring rubric and search strategy for prototype migration

Skill Reference


inn-weekly-feedback-sweep

Scheduled-task wrapper for the weekly Slack feedback sweep. Runs every Monday at 08:00 (or on-demand), produces a structured digest, and posts it to #inn-ai-developers for human triage.

Methodology, channel map, urgency rubric, and digest template are imported from inn-slack-research/resources/feedback-sweep.md so the two skills stay in sync.

Use when you need to:

  • run the weekly Monday feedback sweep
  • regenerate the digest on demand for a custom date window
  • run a catch-up sweep over a longer historical window

Skill Reference


inn-discovery-intake

Interactive discovery interviewer for AI agent build requests. Conducts a structured 10–20 minute conversation, fills the 18-point discovery checklist (stakeholder context, current process, tools, decision points, pain points, edge cases, work examples, SME team), and creates a Wrike Feature Request ready for PM review.

Use when:

  • someone submits an agent build idea or "we should automate X" request
  • a 💡 Feature Request signal from the weekly feedback sweep needs to be formalized
  • a one-line Slack comment needs to be turned into actionable Wrike scope before it enters the backlog

Wrike destination: "6. AI & Innovation Feedback & Feature Requests" folder (IEAFA4LKI5DWLUPH), Feature Request item type (IEAFA4LKPIADDPSI).

Author: Lydia Marasá-Scafidi.

Skill Reference


The Slack Feedback Loop

inn-slack-research, inn-weekly-feedback-sweep, and inn-discovery-intake work together as a passive feedback intake system.

The problem: feedback on skills and agents happens organically in Slack — someone posts "this skill saved me 2 hours" in #artificial-intelligence-gpt, or a pilot rollup in #inn-leadership notes "every Paid Agent entry mentioned timeouts." That signal disappears. The Wrike feedback form exists but people don't use it; they share reactions in the flow of work.

The system captures that passive feedback automatically, structures it for human review, and routes it into Wrike (and eventually GitHub) so the innovation team can act on it.

Slack (organic feedback)
  Weekly Sweep                ← inn-weekly-feedback-sweep (scheduled Mon 08:00)
  Tier 1/2/3 channel search   ← inn-slack-research (methodology + on-demand mode)
  Classify Signals             Bug · Feature Request · Positive · Gap · Adoption
  Weekly Digest                Posted to #inn-ai-developers
  Human Triage                 Review digest, approve routing
   ┌────┴───────────────────────────────┐
   ▼                                    ▼
🐛 Bug → Wrike ticket               💡 Feature Request
❓ Gap → Guru / skill update        → inn-discovery-intake interview
✅ Positive → Release notes         → Wrike Feature Request
📈 Adoption → Ops metrics
  Agent-driven updates         ← inn-seer-agent-skill-developer
  (future state)

Channel map

Tier Channel Why
1 #artificial-intelligence-gpt Primary community AI channel — highest signal density
1 #inn-ai-developers Builder channel — bug reports, architecture questions
1 #help-tips-claude Help questions = implicit friction/gap signals
1 #chatgpt-claude-migration Now a primary adoption/friction channel as workflows migrate
2 #inn-leadership Pilot rollup summaries, cross-team patterns
2 #inn-prototype-seer-mcp MCP-specific issues
3 #seo-team-leads, #tseo-team, #geo-team, #extended-marketing-team Skills used in the wild

Signal taxonomy

Type Emoji Route to
Bug / Issue 🐛 Wrike bug ticket
Feature Request 💡 inn-discovery-intake → Wrike Feature Request
Positive Signal Release notes / internal comms
Question / Gap Guru card or skill update
Adoption Observation 📈 Ops usage tracking

Future state

  1. Now: sweep → digest → human triage → manual Wrike routing
  2. Next: digest → human approves routing actions → agent creates Wrike tickets automatically
  3. Later: triaged bugs and gaps feed directly into inn-seer-agent-skill-developer to propose skill patches for human review
  4. Vision: a dedicated #inn-ai-feedback Slack channel where agents both post digests and respond to new feedback in real time, closing the loop without human routing for low-risk items

Ownership

Component Owner
inn-slack-research skill Jordan Strauss
inn-weekly-feedback-sweep skill Jordan Strauss
inn-discovery-intake skill Lydia Marasá-Scafidi
Weekly digest review Innovation team
Wrike routing (bugs) Jordan Strauss
Wrike routing (feature requests) Lydia Marasá-Scafidi (PM)
Guru / documentation gaps Skill owners
Ops metrics (adoption) Natasha Taylor

Operating Focus

Innovation now spans four high-leverage capabilities:

  1. Build quality (inn-seer-agent-skill-developer)
  2. Research quality (inn-slack-research)
  3. Continuous feedback intake (inn-weekly-feedback-sweep)
  4. Structured discovery (inn-discovery-intake)

This keeps the division focused on practical acceleration for release and platform maturity — and on closing the loop from real user signal back into agent and skill improvements.


Last updated: 2026-04-29