Hi Andrew and Justin,

I find this experiment incredibly useful! While human verification is key
to catching any potential hallucinations, this structure offers a good
foundation for reviewers to validate project claims.

To make this even better, I suggest adding a "Human-In-The-Loop" section
with direct links to the data sources. Regarding the "development activity"
point, the data actually shows a nice doubling in commit frequency when
comparing the last three months to the yearly total. Clickable links would
clarify this; Andrew, perhaps this was the source of your confusion, which
the right "source links" could resolve?

Regarding the environmental and financial side - very valid cocerns indeed
- I also was very sceptical even half a year ago, but the situation - even
during those 6 months - changed dramatically for the better. And the mcp
tool calls are quite useful for illustrating this:

  - Cost: Roughly $0.30–$1.00 in API spend.
  - Water: About 0.1–0.2 liters for cooling.

This usage is comparable to a cup of coffee and is likely very efficient
compared to the "energy/water" used when writing from scratch and spending
more time on it by human. The verbosity Andrew mentioned can also be easily
adjusted via settings to match everyone’s preference for conciseness; this
is another possible area for improvement.

In projects like Magpie, we are also exploring local LLMs to further reduce
costs and stay vendor-neutral.

Also there are initiatives/startups (even run by my friends - ASF members)
to get ASICs on the market that are supposed to drive the cost / energy of
LLM operations several orders of magnitude down. So, unlike the cost and
energy of a cup of coffee, this is likely only trending heavily downwards
(there are reports showing 30x cost/energy drop over last 2 years  on
average per token - and it shows no sign of slowing down).

Best,
Jarek


On Sat, May 30, 2026 at 11:26 AM Andrew Wetmore <[email protected]> wrote:

> The statement "Development activity has increased over the reporting
> window" does not match the data provided.
>
> This is an interesting exercise, but is the result not more verbose
> (drowning the reader in details) than required for a podling report? How
> many credits, how much fresh water, were spent in its creation?
>
> A
>
> Andrew Wetmore
> Assistant VP, Marketing and Publicity, The ASF <https://apache.org>
> Editor-Writer, Infra team, The ASF
>
> Editor, moosehousepress.com
>
> On Sat, May 30, 2026, 4:39 AM Justin Mclean <[email protected]>
> wrote:
>
> > Hi,
> >
> > The recent MCP discussed obtaining most of the data needed to create a
> > draft report for a project.
> >
> > What would people think about populating the incubator report with a
> draft
> > report along these lines?
> >
> > Here’s one I generated for Hamilton (I’ve not edited this in any way). I
> > know the Hamilton report will be different from this, but as you can see,
> > it can generate useful information in a few minutes.
> > Apache Hamilton (Incubating) -- May 2026 Report
> >
> > Hamilton is a lightweight in-process framework to define, execute, and
> > observe directed acyclic graphs (DAGs) that express data transformations.
> > In Hamilton, one can express complex DAGs of transformations, e.g. from
> > dataframe transformations (using pandas, polars, PySpark), machine
> learning
> > pipelines, through to regular software engineering API request and LLM
> API
> > based workflows. Observability hooks are built into the framework. The
> > Hamilton UI is a self-hostable service to capture observability output
> from
> > workflow runs.
> >
> > Hamilton has been incubating since 12 April 2025.
> >
> > Three most important unfinished issues to address before graduating:
> >
> > Release cadence for the core package. The last ASF-approved source
> release
> > on dist.apache.org is 1.89.0-incubating (October 2025). Vote threads are
> > currently open for sdk, ui, lsp, and contrib components, but a new ASF
> > release of the core hamilton package is overdue. The project should
> > complete at least one further approved release before graduation.
> >
> > PyPI package naming and provenance. The sf-hamilton package on PyPI
> > (version 1.90.0) does not carry the ALv2 license field or an Incubator
> > disclaimer in its project description. It is also unclear whether all
> > versions published to sf-hamilton and apache-hamilton were produced
> solely
> > from IPMC-approved ASF releases; this should be confirmed and documented
> > before graduation.
> >
> > Mentor engagement. The average mentor sign-off rate over the past 12
> > months is 1.6 per report, which is below the level expected for
> consistent
> > oversight. Mentors should confirm they are actively reviewing each report
> > before graduation discussions proceed.
> >
> > Are there any issues that the IPMC or ASF Board need to be aware of?
> >
> > No.
> >
> > How has the community developed since the last report?
> >
> > Development activity has increased over the reporting window. Over the
> > past three months the project recorded 128 commits from 16 unique
> > committers, compared with 257 commits from 27 committers over the full
> > 12-month window -- indicating that activity is concentrated in a smaller
> > group recently but has not dropped off. The bus factor (50th percentile)
> > has held at 2 across all windows, meaning the project remains dependent
> on
> > a small core of contributors; the 75th percentile is 3-4, which is an
> > improvement over the year.
> >
> > New contributor attraction remains healthy: 12 new contributors appeared
> > in the last three months, consistent with the 12-month trend of 21.
> Mailing
> > list participation on dev@ has been moderate: around 40 messages from 9
> > unique posters in the past three months. That is lower than the 12-month
> > figure of 151 messages from 13 posters, suggesting the pace of general
> > discussion has softened outside of release activity.
> >
> > A proposal for an Airflow Provider for Hamilton was posted to dev@ in
> > April 2026 and attracted community discussion, which is a positive sign
> of
> > broadening ecosystem interest.
> >
> > How has the project developed since the last report?
> >
> > The project is actively preparing multiple component releases. As of the
> > reporting date, vote threads are open on dev@ for four sub-packages:
> > apache-hamilton-sdk 0.9.0, apache-hamilton-ui 0.0.18, apache-hamilton-lsp
> > 0.2.0, and apache-hamilton-contrib 0.0.9 (all RC0, incubating). These
> votes
> > were opened on 26 May 2026.
> >
> > The most recent completed ASF source release is 1.89.0-incubating, dated
> > 11 October 2025, and is available at archive.apache.org with signature
> > and SHA-512 checksum present. A GitHub release tagged
> > apache-hamilton-v1.90.0-incubating-RC0 was published on 25 April 2026 but
> > is flagged as a release candidate and has not yet completed the formal
> vote
> > process.
> >
> > The apache-hamilton package on PyPI is at version 1.90.0 and includes an
> > Incubator disclaimer in its description. The older sf-hamilton package
> > redirects installers to apache-hamilton but its PyPI metadata does not
> > include the ALv2 license field or an Incubator disclaimer.
> >
> > How would you assess the podling's maturity?
> >
> > [ ] Initial setup
> > [ ] Working towards first release
> > [ ] Community building
> > [x] Nearing graduation
> > [ ] Other
> > Date of last release:
> >
> > 11 October 2025 (apache-hamilton-1.89.0-incubating). Multiple component
> > release votes (sdk, ui, lsp, contrib) are currently in progress as of 26
> > May 2026.
> >
> > When were the last committers or PPMC members elected?
> >
> > [PODLING TO COMPLETE]
> >
> > Have your mentors been helpful and responsive?
> >
> > Mentors are listed as Kevin Ratnasekera, Ayush Saxena, PJ Fanning, and
> > Jarek Potiuk. Mentor participation is visible on dev@ -- PJ Fanning and
> > Jarek Potiuk have engaged in recent release vote threads. The average
> > mentor sign-off rate over reported periods is 1.6, which is below the
> > expected level; the PPMC should confirm that all four mentors are
> reviewing
> > reports regularly.
> >
> > Is the PPMC managing the podling's brand / trademarks?
> >
> > [PODLING TO COMPLETE]
> >
> >
> >
> > This was the prompt to create it:
> >
> > System prompt: Use only data retrieved from the Incubator MCP, MailMCP,
> > ReleaseMCP, Health MCP, and Policy MCP to populate the report. Do not use
> > any previous report content or training data. Follow these rules:
> >
> > The reporting window runs from the first day of the month of the last
> > report to today. Determine this from the podling's reporting schedule
> > retrieved in step 0.
> > Do not include raw statistics as standalone facts. Instead interpret them
> > as trends by comparing across available time windows (3m, 6m, 12m). Note
> > whether activity is growing, stable, or declining in plain language.
> > Only include a statistic directly if it is meaningful on its own, such as
> > the number of releases, a specific release version, or a release vote in
> > progress.
> > Surface health concerns in the issues section -- such as declining
> > reviewer diversity, concentration of commit activity, or dropping mailing
> > list engagement -- rather than leaving it blank.
> > The "Are there any issues that the IPMC or ASF Board need to be aware
> of?"
> > section is for issues that require board-level attention -- such as legal
> > concerns, governance breakdowns, or matters the IPMC cannot resolve
> alone.
> > Routine release activity, compliance housekeeping, and community health
> > watch items belong in other sections. If there are no board-level issues,
> > write "No."
> > Only call the Policy MCP if ReleaseMCP or Health MCP identifies a
> > potential compliance concern. If called, reference the relevant policy in
> > the issues section.
> > For any section the tools cannot populate, write [PODLING TO COMPLETE] on
> > its own line.
> > Do not copy forward language from any previous report.
> > Write in plain, concise English suitable for an international audience.
> > Avoid jargon.
> > Output the completed report as a markdown document. User prompt: Draft
> the
> > Incubator board report for Apache {podling} (Incubating) using the
> > following steps:
> > Call the Incubator MCP to retrieve the podling's description, incubation
> > start date, and reporting schedule to determine the current reporting
> > window start date.
> > Call MailMCP to retrieve dev list activity since the start of the
> > reporting window, looking for release votes, discussion threads, and
> > significant proposals.
> > Call ReleaseMCP to retrieve current release artifact evidence including
> > dist, archive, PyPI, and GitHub.
> > Call the Health MCP to retrieve community health metrics across all
> > available time windows.
> > If ReleaseMCP or Health MCP identifies a potential compliance concern,
> > call the Policy MCP to confirm the relevant policy and reference it in
> the
> > issues section.
> > Synthesise the results into the standard Incubator report template below
> > and output as markdown. Template: markdown
> > ## Apache {podling} (Incubating) -- {report month and year} Report
> >
> > {description from Incubator MCP}
> >
> > {podling} has been incubating since {start date from Incubator MCP}.
> >
> > ### Three most important unfinished issues to address before graduating:
> >
> > ### Are there any issues that the IPMC or ASF Board need to be aware of?
> >
> > ### How has the community developed since the last report?
> >
> > ### How has the project developed since the last report?
> >
> > ### How would you assess the podling's maturity?
> > - [ ] Initial setup
> > - [ ] Working towards first release
> > - [ ] Community building
> > - [ ] Nearing graduation
> > - [ ] Other
> >
> > ### Date of last release:
> >
> > ### When were the last committers or PPMC members elected?
> >
> > ### Have your mentors been helpful and responsive?
> >
> > ### Is the PPMC managing the podling's brand / trademarks?
> >
> > And this is the list of MCP calls it made:
> > Here are all MCP tool calls made, in order:
> >
> > podlings:reporting_schedule (name="Hamilton") -- retrieved Hamilton's
> > reporting group, cadence, and current report period to establish the
> > reporting window.
> > ipmc:podling_brief (podling="Hamilton", brief_format="detailed") --
> > retrieved incubation start date, mentor list, health summary, and key
> > indicators including commit counts, committer diversity, and active
> > concerns.
> > ipmc:release_artifact_evidence (podling="Hamilton",
> > github_project="hamilton", include_platforms=true,
> > pypi_packages=["sf-hamilton", "apache-hamilton"]) -- retrieved
> dist/archive
> > artifact evidence, GitHub releases, PyPI package metadata, Docker Hub
> > checks, and Maven checks.
> > apache-health:compare_windows (podling="Hamilton", windows=["3m", "6m",
> > "12m"]) -- retrieved community health metrics across all three windows
> for
> > trend comparison (commits, committers, authors, dev posters, PRs,
> releases,
> > mentor sign-offs, bus factor, reviewer diversity).
> > incubator-mail:recent_podling_mail (podling="hamilton", list_name="dev",
> > timespan="lte=3M", limit=20) -- retrieved recent dev@ activity including
> > the four active release vote threads (sdk, ui, lsp, contrib) and the
> > Airflow Provider proposal discussion.
> > podlings:get_podling (name="Hamilton") -- retrieved the full podling
> > description, champion, mentor names, and start date for inclusion in the
> > report header.
> > Kind Regards,
> > Justin
>

Reply via email to