start using Elasticsearch with your agents
sign up, open a project, and add Elasticsearch to its setup script in settings → server.
search bugs are almost never in the query. they are in the mapping, the analyzer, or the document that was indexed differently from the one being searched for. installing elasticsearch in each conversation's sandbox lets an agent settle which of those it is by trying it.
what agents do with it
once the setup script has run, elasticsearch is running in the conversation's sandbox and reachable over http from the same box. agents create an index with the app's mapping, load representative documents, run the query, and read the score explanation when the wrong document comes back first. tuning is then a loop rather than a guess: change the analyzer, reindex, search again.
working from real data
a document export the agent downloads lands in the conversation's sandbox, so an agent can index the actual export rather than a made-up sample. because projects are isolated, that data stays inside the project that fetched it, and no other project can search it.
example tasks
- 01
searching for a hyphenated product code returns nothing. work out why.
the agent indexes the document on the project's elasticsearch, runs the analyzer over the field, sees how the code was tokenised, and adjusts the mapping until the term matches.
bashreadedit - 02
the wrong article ranks first for our most common query. fix the relevance.
it loads the real corpus, runs the query with an explain, and reports which field boost is doing the damage before changing anything.
bashgrep
what to know before installing
the project needs a sandbox, and this one is a genuine reason to pick the standard size or larger. a setup script for it adds the elastic apt repository, installs the package, and enables the service, and settings → server shows that run's output.
how to install Elasticsearch
Elasticsearch goes in the project's setup script, not once per team and not on your laptop.
- give the project a sandbox
Elasticsearch runs in each conversation's sandbox, on your team's own sandbox provider account. an owner or admin connects a provider once on the team's providers page, then picks it and a default size for the project, at creation or later in settings → server. a project with no sandbox has nowhere to run it.
- add it to the setup script
settings → server holds the project's setup script, a bash script that runs once on each new sandbox, in /project, with sudo. add Elasticsearch's install commands there and save.
- the next sandbox runs it
the script runs when a conversation's sandbox is created or reused, on the conversation's first turn. settings shows the last run with its exit code and output, and a failed run leaves the sandbox usable rather than blocking the project.
- agents use it through bash
from then on every agent in the project reaches Elasticsearch with the bash tool, on the same filesystem as the checkout they are working in. the conversation's sandbox keeps it between turns, sleeping when idle and waking with the next one.
where Elasticsearch actually lives
each conversation in a project gets its own sandbox on your team's sandbox provider account, and the project's setup script in settings → server decides what is on it. two projects never share a filesystem, so a version pinned for one client cannot collide with another's.
the sandbox provider bills its machine time to your team's account while it runs, and aldena adds nothing. the sizes are on isolated projects, and what agents may run in the sandbox is governed by the project's tool permissions.
faq
what does installing elasticsearch on a project give the agents?
a running elasticsearch service in each conversation's own sandbox. agents create indices, load documents, and run search queries over http from the same box, so relevance work happens against the real analyzers instead of a description of them.
why does search work need a live instance?
because analyzers and mappings decide the result, and neither is obvious from reading the code. an agent that can index a document and search for it sees why a term did not match, which is the only way to tell a mapping problem from a query problem.
how much memory does it need?
elasticsearch is the heaviest technology in the catalog. a project running it alongside an application wants the standard size or larger. your sandbox provider bills the machine time to your team's account, and aldena adds nothing.
does the index survive between sessions?
only within a conversation. each conversation's sandbox keeps its files between turns, sleeping when idle and waking with the next one, so an index built today is still there tomorrow. a new conversation gets its own sandbox, and the setup script runs on it again. a conversation left quiet for 7 days gives its sandbox back, and its next turn starts fresh.
spin up your first project.
one project per client, product, or piece of work, staffed with intake, requirements, build and code review agents.
