me.freehirePublished by freehire.meOfficial publisherv0.8.0Updated Oct 30, 2026
freehire.me: search 3.3M+ IT jobs from 294K company boards, track applications, tailor CVs.
Searches freehire's IT job catalogue, inspects postings and employers, and reports salary and market statistics.
Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.
Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.
Read https://sourceweft.com/mcp/me.freehire%2Ffreehire/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.Add this to your client's mcpServers config.
{
"mcpServers": {
"freehire": {
"type": "http",
"url": "https://freehire.me/api/v1/mcp"
}
}
}Search 4.9M+ unique open jobs from 437 sources inside Claude, ChatGPT or your coding assistant. Find roles by stack, salary, seniority, work mode and location, inspect employers, and keep your shortlist and application progress in one place. The catalogue includes 680K+ company profiles and 178 ATS platforms. Figures are from the exact catalogue snapshot on 5 November 2026 (live figures).
The hosted MCP connection offers public search without an account and personal tools through OAuth at one address. The local npm server uses a personal API key and also provides CV and experience-bank tools. Saving or marking a job applied updates tracking; employer submission is a separate OAuth workflow that requires explicit permission and review.
The local server mirrors the freehire CLI: same API, same credentials, exposed as MCP tools instead of shell commands.
The plugin connects to https://freehire.me/api/v1/mcp over Streamable HTTP.
Public catalogue tools work without sign-in. Connect your account with OAuth for personal tools. It includes job
search and market research, plus your profile, saved jobs, application tracking,
reports, alerts and account-enabled automatic applications. In Claude Code, install
it as a plugin:
In other OAuth-capable hosts, add the canonical URL as a custom remote connector and
complete the sign-in flow. In Claude Code, use /mcp to select Freehire and authenticate
before personal calls; for a custom server named freehire, use claude mcp login freehire.
Do not assume every host opens sign-in automatically from a tool-level challenge. The included skill automatically submits the user's original
unrecognized skill, specialization or company industry for dictionary review through
suggest_dictionary_term. It also sends a minimal report_problem for a failed operation
or missing capability, excluding credentials, CV content, conversations and personal
contact details. These two workflow reports require no separate confirmation; empty
search results alone trigger neither. The agent confirms submission only after success
and continues the original task. Other account changes and application sending require
your confirmation; automatic applications also require the account setting that permits
connected agents to apply.
The ChatGPT package is freehire (display name Freehire), with a freehire
MCP connection to the original URL and mixed public/OAuth authentication. Upload
updates to the existing Freehire plugin. The separate Freehire Account plugin and
its /api/v1/mcp/account endpoint remain available during client migration.
Every push to main, pull request, and manual run of the
CI workflow
builds two ZIP files after the checks pass. Open a successful run and download
the freehire-plugins-<commit> artifact:
freehire-chatgpt-<version>.zip: upload this ZIP as the ChatGPT plugin package.
It includes the portable plugin manifest, hosted MCP connection, icon, and the
current find-tech-jobs skill.freehire-claude-skill-<version>.zip: upload this ZIP in claude.ai under
Customize → Skills. Add https://freehire.me/api/v1/mcp as a remote
connector separately and complete OAuth sign-in; uploading a skill does not
configure an MCP connection.Run npm run build:plugins locally to write both archives to dist/plugins/.
Packaging uses Python 3.9+ and its standard library; it needs no API credentials,
live server access, or sibling checkout. The version comes from package.json,
and all packaged metadata, assets and skills come from the checked-out commit.
Both archives include the same skills/find-tech-jobs/SKILL.md source and its dictionary
and problem-reporting workflow. The hosted MCP connection discovers the live server's
tools rather than bundling a fixed tool list; existing account tools remain available.
The project-local .agents/skills/verify-freehire-mcp skill verifies the archives,
local stdio discovery, hosted public access and OAuth using real transports and
the official MCP Inspector CLI. CI runs the archive/local checks and uploads the
freehire-verification-<commit> proof artifact. Hosted checks additionally need
the deployed unified backend and a dedicated reviewer OAuth grant.
This workflow builds download artifacts; directory submission and publication remain separate steps.
For the public Claude directory, submit this repository with Plugin path set
to plugins/claude and track main. Pair the bundle with the existing Freehire
MCP connector in the same Claude organization. The bundle includes the skill and
the same hosted MCP URL, so it does not install the local npm server or its
dependencies. The existing connector remains the source of authentication and
tool health information.
npm run build:plugins also refreshes the committed plugins/claude folder from
the current skill, Claude manifest, connection configuration, README template
and license. Commit that folder with changes to those sources. CI checks it
before building the ZIPs and fails if the directory bundle is stale. After the
initial directory review, Claude can scan new commits on the tracked branch;
publishing still follows the listing's review settings.
The hosted tools use names such as search_jobs, get_job, get_profile,
save_job, track_job and my_jobs; discover their current schemas in your host.
The unified hosted search uses top-level public filter fields; discover its live schema.
Personal tools may use nested filters, so inspect each tool separately. CV tailoring and experience
bank tools in the local server below are not part of this account connection.
No global install needed — the host runs it via npx. Add it to your host's MCP
configuration (Claude Desktop → Settings → Developer → Edit config, or
~/.claude.json for Claude Code):
Create the fhk_… key in the web app (freehire.me → account menu → API keys).
If you already use the freehire CLI (freehire auth login), you can omit env —
the server reads the same ~/.freehire/creds.json.
The token and API base URL resolve with precedence
env → ~/.freehire/creds.json → default https://freehire.me:
The server only reads the credentials file (it never writes it — logging in stays the CLI's job). If no token is configured, tools return a clear "not authenticated" error rather than the server failing to start.
Filters. search, market_fit, and facets share the same market-filter
parameters: remote, region, country, city, company, category, role,
seniority, employment_type, english_level, exclude_skill, salary_min, visa,
plus a generic facets map ({"source": "greenhouse"}) for any other facet in the
vocabulary. Discover valid values with the facets tool — do not invent them. In
search, skills is a filter; in market_fit, skills is the measured set.
Geography widens. region, country and city are ONE OR-group: region: ["eu"]
with country: ["IT"] means "in Europe or in Italy" and returns everything the
region alone would. To search a single country, pass country and omit region. The
three name a single concept — where — so picking two places reads as "either", which
is what makes region: ["eu"] with country: ["BR"] ("Europe or Brazil") useful. There
is no AND to switch on: _mode=and does not apply to geography.
Unknown values are named, not guessed. A filter value the API does not recognise (a
misspelt skill such as golangg) is reported in ignored_values, e.g.
{"skills": ["golangg"]}, beside total in search and in the {data, ignored_values}
wrapper of facets and market_fit. Unlike an unread param it still narrows — to nothing —
so a 0 beside it is a typo, not an empty market. Skill aliases resolve server-side
(golang searches go); list_filter_values names the real value.
Unread params are ignored, not refused. A filter key the API does not recognize
does not fail the request, it widens it. Such keys come back in the result's ignored
list, with did_you_mean when only the grammatical number was wrong. search reports it
alongside total; facets and market_fit answer a single object, so they wrap it as
{data, ignored} — and only then, leaving a clean call's shape untouched. Any number from
a result carrying ignored answers a broader question than the one asked — retry with the
suggested name before reporting it.
Forwarded tools. check_jobs_open, similar_jobs, salary_stats,
job_market_stats and list_filter_values have no REST route, so they are forwarded to
the unified hosted MCP endpoint (/api/v1/mcp, no key sent for public reads). The five
job-alert tools use the same endpoint with your full-access API key, so an alert made
here is exactly the website's — the same saved search, channel rule and match ledger.
Their answers arrive in the hosted server's own words. Automatic applications
(auto_apply_*) require OAuth on that same endpoint and are not offered to an API key.
Searching well. Join a role's two or three names with an uppercase OR
(title:SRE OR title:"site reliability"), put location, work mode, seniority and skills in
filters rather than the query, loosen one constraint at a time when few results come back,
shortlist from the search results and run check_jobs_open before presenting.
Descriptions. search reads the API's agent endpoint, so every hit already carries
the posting's full description rendered as markdown — a host can screen a result set
without a job call per hit. Descriptions are long, so keep limit modest.
The evidence rule. Every achievement in the bank records who asserted it.
cv_import, stated_in_chat and manual mean the candidate did, and may be cited on a
CV; agent_inferred means a model read it into the record, and may not. cv_edit
refuses any claim about the candidate without an evidence_id pointing at a citable one,
which is why experience_list is the tool that makes cv_edit usable at all.
Correcting an achievement does not move that label: an agent_inferred one stays
uncitable however it is reworded. The only way it becomes citable is to ask the
candidate, then record what they say with experience_add_achievement.
Removing is final — the bank has no undo. A place must be emptied before it can go, because deleting one would take every achievement under it. Folding two achievements into one, keeping the numbers from both, is on the site.
Each tool returns the raw API data as JSON text; an API error becomes an isError
result carrying the HTTP status (a 401 adds an auth hint).
MIT — see LICENSE. The freehire backend and CLI are MIT too.
Source: README.md at commit eea1829
Check whether postings are still open
Check whether postings are still open, by the slugs search results carry — up to 50 per call. Use before acting on a saved shortlist: each posting answers open, closed or not_found, with when it closed or was last seen live.
slugsrequiredRead one job posting (research)
Read one posting in full by an id search returned: its full text, the employer, the board it came from and whether it still looks live.
idrequiredRead one employer
Read what freehire holds about one employer by the slug a result carries: what they do, which industries they are in, where they are based, and how many postings they have open.
slugrequiredRead one job posting
Read one posting in full by the slug a search result carries: the whole description, the stated requirements, which applicant tracking system handles applications, and whether the posting still looks live — when it was last seen open, a reality class, and any observed signs that it is not being filled.
slugrequiredJob market breakdown
How many open postings match a set of filters, and their most common role families, seniorities, skills, work modes, countries and regions. Takes the same filters as a job search.
limitquerycitiesoffsetskillsregionssourcescategorysemanticcountriessenioritywork_mode+11 moreList filter values
The values one job-search filter accepts — skills, categories, seniorities, sources, countries and the rest — each with how many open postings carry it, most common first. A value not listed matches no posting. Use it before filtering on a value you are not sure of.
limitfilterrequiredcontainsSalary ranges
Salary percentiles (25th, median, 75th) from open postings that state pay, scoped by role family, seniority and country, with the sample size behind each figure. Currencies and pay periods are reported separately, never converted.
countrycategorysenioritySearch jobs (research)
Search freehire's IT job postings by free text. Returns matching postings as {id, title, url}.
queryrequiredSearch employers
Find employers in freehire's catalogue by name or by where they hire, with how many postings each has open. Use when the question is about a company rather than about a role.
limitqueryoffsetcountriesSearch jobs
Search freehire's catalogue of IT job postings. Use for any question about which jobs are open — by technology, seniority, location, pay or employer. Returns a short preview of each posting. A role has several names: join two or three in one query with OR, e.g. title:SRE OR title:"site reliability". Location, work mode, seniority and skills go in their filters, never in the query. If suspiciously few results come back, loosen one constraint at a time and say which. Present results as a table of title, company, location, posted date and link (the employer's posting and ours), naming the source. Set semantic for a need described in words ("keep our payment systems secure"), not for a job title or a non-technical role, where plain keywords answer better.
limitquerycitiesoffsetskillsregionssourcescategorysemanticcountriessenioritywork_mode+11 moreFind similar postings
Postings most similar to one posting, by its slug, nearest first — for "show me more like this". Empty for the newest postings, whose neighbours are not computed yet.
slugrequiredlimit100% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
/plugin install freehire --marketplace strelov1/freehire-mcp{ "mcpServers": { "freehire": { "command": "npx", "args": ["-y", "freehire-mcp"], "env": { "FREEHIRE_TOKEN": "fhk_xxxxxxxx" } } }}npm installnpm test # vitest: config, client (mock server), facets, tool dispatchnpm run build # tsc → dist/