Merge pull request #78 from fullsend-ai/chore/update-customer-research-skill

chore: update customer-research skill for June 2026

Author: ralphbean

customized/skills/customer-research/SKILL.md

@@ -20,43 +20,54 @@ prioritization or customer impact.
 ## Project status
 
 fullsend-ai passed its MVP milestone on April 23, 2026. The project is
-early-stage and actively onboarding its first users.
+in active adoption, executing against revised 90-day goals (target: June
+30, 2026) focused on bug-fix workflow adoption, feature refinement
+capabilities, and evaluating trustworthiness of agent behaviors.
+
+A key near-term focus is multi-platform support (especially GitLab),
+driven by demand from multiple customers and use cases.
+
+All agentic workflows are hybrid — teams use fullsend alongside manual
+and locally-driven methods. Auto-merge is not an expectation at this
+phase; agent-produced code is reviewed by humans before merging.
 
 > **Staleness warning:** The customer details below are a point-in-time
-> snapshot. Where possible, commands are provided to fetch live data.
-> Static content should be periodically reviewed and updated.
+> snapshot (last updated June 2026). Where possible, commands are
+> provided to fetch live data. Static content should be periodically
+> reviewed and updated.
 
 ## Strategic customers
 
-There are three strategic external customers listed below. The
-fullsend-ai org itself is also a user (dogfooding), but the external
-customers are the ones that matter for prioritization.
+The strategic customers are listed below. The fullsend-ai org itself is
+also a user (dogfooding), but the external customers are the ones that
+matter for prioritization.
 
-All three are high priority. Issues and feedback from these customers
-are direct signals of the new-user onboarding experience and should be
-treated with urgency. Other users are welcome and should be supported,
-but these three take precedence when prioritizing work.
+Issues and feedback from these customers are direct signals of the
+onboarding experience and should be treated with urgency. Other users
+are welcome and should be supported, but strategic customers take
+precedence when prioritizing work.
 
 ### 1. konflux-ci
 
-Several repositories in the `konflux-ci` GitHub org are onboarded. To
-get the current list of enrolled repositories, run:
+Many repositories in the `konflux-ci` GitHub org are onboarded — the
+number has expanded significantly since initial onboarding. To get the
+current list of enrolled repositories, run:
 
 ```bash
 gh api repos/konflux-ci/.fullsend/contents/config.yaml \
   --jq '.content' | base64 -d | yq .
 ```
 
-We expect a few dozen more konflux-ci repositories to onboard during
-May 2026.
+The 90-day goal is to continue scaling usage across konflux-ci and to
+incorporate feature refinement into the existing process.
 
-### 2. openkaiden (via @deboer-tim)
+### 2. redhat-developer (via @deboer-tim)
 
-@deboer-tim is evaluating fullsend-ai for the openkaiden project. He
-created a personal fork org —
-[openkaiden-fullsend](https://github.com/openkaiden-fullsend) — so he
-can test the integration without disrupting the real openkaiden org's
-development workflow while fullsend-ai is still young.
+@deboer-tim is the primary contact for the Developer teams. The goal
+has shifted from openkaiden to onboarding 2 or more repos from the
+`redhat-developer` GitHub org. He previously used a personal fork org —
+[openkaiden-fullsend](https://github.com/openkaiden-fullsend) — for
+early testing.
 
 Issues he has filed:
 [fullsend-ai/fullsend issues by deboer-tim](https://github.com/fullsend-ai/fullsend/issues?q=is%3Aissue+author%3Adeboer-tim).
@@ -71,9 +82,29 @@ demonstrate to other guacsec maintainers what the workflow looks like
 and whether the platform should be considered safe. Trust and
 transparency are key concerns for this customer.
 
-As of April 2026, guacsec has not yet onboarded (no `.fullsend` repo
-exists). @mrizzi has not filed issues directly but is mentioned in
-issues [#457](https://github.com/fullsend-ai/fullsend/issues/457) and
+Onboarding guacsec upstream is a 90-day goal (target: June 30, 2026).
+The plan is for fullsend to be installed in at least one guacsec repo,
+with Red Hat P&D paying for inference, and upstream maintainers agreeing
+to try it.
+
+@mrizzi has not filed issues directly but is mentioned in issues
+[#457](https://github.com/fullsend-ai/fullsend/issues/457) and
 [#459](https://github.com/fullsend-ai/fullsend/issues/459), which
 relate to local execution and pre-adoption evaluation — likely driven
 by his need to demo the platform safely.
+
+### 4. RHDH / Developer teams (via @durandom)
+
+@durandom and the RHDH (Red Hat Developer Hub) team have started using
+fullsend. The Developer teams are a 90-day goal for bug-fix workflow
+adoption, though onboarding is currently blocked on reducing install
+overhead and simplifying customization options. Unblocking these teams
+is a priority for the fullsend team in June 2026.
+
+### 5. fullsend-playground (new interest via @ascerra's demo)
+
+@ascerra ran a demo on June 3, 2026 inviting people to try fullsend
+via the `fullsend-playground` GitHub org. This has generated a large
+number of new interested users checking out the project. These are
+early explorers — not yet strategic customers — but their feedback is
+valuable signal on the out-of-box experience and should be monitored.