pickfit

BrowseDeveloper CLIs

mi-lsp

Local CLI that indexes repos to find code, read exact ranges, and follow canonical docs.

Tested on macOS and Linux

Use it

macOS arm64

  1. download install.sh to read before running

    curl -fsSL https://raw.githubusercontent.com/fgpaz/mi-lsp/main/scripts/install/install.sh -o /tmp/install.sh && echo DONE && wc -c /tmp/install.sh
  2. download sibling validator required by install.sh

    curl -fsSL https://raw.githubusercontent.com/fgpaz/mi-lsp/main/scripts/install/validate-worker-manifest.js -o /tmp/validate-worker-manifest.js && echo DONE && wc -c /tmp/validate-worker-manifest.js && ls /tmp/*.sh /tmp/*.js
  3. run CLI-only install.sh for darwin-arm64

    sh /tmp/install.sh
Where the README and reality differ
  • README: curl -fsSL .../install-agent.sh | sh installs directly
    What worked: download validate-worker-manifest.js next to install.sh before running install-agent.sh/install.sh (curl .../validate-worker-manifest.js -o /tmp/validate-worker-manifest.js)

Check that it works

mi-lsp --help

The output should contain “version for mi-lsp”.

First thing to try

mkdir -p src/billing tests/billing .docs/wiki/05_requirements && cat > src/billing/retry.py <<'EOF'
"""Billing retry logic sample."""
RETRY_LIMIT = 3

def billing_retry(order_id, attempts=RETRY_LIMIT):
    """Retry billing for order."""
    for i in range(attempts):
        if charge(order_id):
            return True
    return False

def charge(order_id):
    """Charge order."""
    return True
EOF
cat > tests/billing/retry_test.py <<'EOF'
from src.billing.retry import billing_retry
def test_billing_retry():
    assert billing_retry("order-123") is True
EOF
cat > .docs/wiki/05_requirements/RF-GPH-010.md <<'EOF'
# RF-GPH-010 - Billing retry spec
Billing retry governs how the daemon restarts failed charges.
See billing_retry symbol.
EOF
echo SETUP_DONE && ls -R src tests .docs
mi-lsp init . --name myapp --format toon

In your repository folder, run mi-lsp init . --name <workspace> --format toon to index it as a workspace.

Linux x86_64

  1. download install.sh to read before running

    curl -fsSL https://raw.githubusercontent.com/fgpaz/mi-lsp/main/scripts/install/install.sh -o /tmp/install.sh
  2. download sibling validator required beside install.sh

    curl -fsSL https://raw.githubusercontent.com/fgpaz/mi-lsp/main/scripts/install/validate-worker-manifest.js -o /tmp/validate-worker-manifest.js
  3. install CLI with --skip-worker-install

    sh /tmp/install.sh --skip-worker-install
Where the README and reality differ
  • README: curl -fsSL .../install-agent.sh | sh (which fetches install.sh alone)
    What worked: download https://raw.githubusercontent.com/fgpaz/mi-lsp/main/scripts/install/validate-worker-manifest.js beside install.sh before running it; install.sh calls node <script_dir>/validate-worker-manifest.js
  • README: mi-lsp nav search "billing retry" --include-content --workspace myapp
    What worked: on a container workspace add --repo src: mi-lsp nav search ... --workspace myapp --repo src --include-content

Check that it works

mi-lsp init . --name myapp --format toon 2>&1

The output should contain “ok: true”.

First thing to try

mkdir -p src .docs/wiki/05_requirements && cat > src/app.py <<'PY'
def hello_world(name):
    """Greet by name."""
    return f"hello {name}"

def billing_retry(attempt):
    """Retry billing."""
    return attempt + 1
PY
cat src/app.py && mkdir -p myapp 2>/dev/null; echo setup-done
mi-lsp init . --name myapp --format toon 2>&1

In your repository folder, run mi-lsp init . --name <workspace> --format toon to index it as a workspace.

What it does

  • Find a symbol by name
    How to use it
    1. Index your repo with mi-lsp init . --name <workspace> --format toon.
    2. Run mi-lsp nav find <symbol> --workspace <workspace> --format toon to get the file, line and definition.
    Tested
  • Search text and see matching code inline
    How to use it
    1. Index your repo with mi-lsp init . --name <workspace> --format toon.
    2. Run mi-lsp nav search "<your text>" --workspace <workspace> --repo <repo> --include-content --format toon to see matches with their code lines.
    Tested
  • Read exact line ranges from one or more files
    How to use it
    1. Index your repo with mi-lsp init . --name <workspace> --format toon.
    2. Run mi-lsp nav multi-read <file>:<start>-<end> <file>:<start>-<end> --workspace <workspace> --format toon to print those lines.
    Tested
  • Browse the project wiki and governance docs
    Not tested

How far pickfit got

macOS arm64

  1. Installs
  2. Starts
  3. 3 of 4 features work
  4. Works offline

Linux x86_64

  1. Installs
  2. Starts
  3. 3 of 4 features work
  4. Works offline

Windows: the project says it runs there; pickfit hasn’t tried it.

Good for

  • Find a symbol by name
  • Search text and see matching code inline
  • Read one or more exact file ranges
  • See what a change affects (callers, tests, docs)
  • Navigate the wiki / governance docs

How pickfit tested it

Each run starts from a clean machine and follows the steps above. How pickfit verifies

macOS arm64

Its main function works

  • Replayed from scratch
  • Works offline
  • Checked on 30 Oct 2026
  • Version 0.16.7

Differences the test machine hit

  • README: install-agent.sh adds CLI plus mi-lsp skill
    In the test: CLI installs fine but the trailing `npx skills add fgpaz/claude-skills` step fails with authentication failure for that repo; CLI-only install.sh avoids it
  • README: mi-lsp nav search "billing retry" --include-content --workspace myapp
    In the test: on a container workspace add --repo <repo> to get inline content instead of a scope preview

Linux x86_64

Its main function works

  • Replayed from scratch
  • Works offline
  • Checked on 30 Oct 2026
  • Version v0.16.7

Differences the test machine hit

  • README: curl -fsSL .../install-agent.sh | sh installs CLI
    In the test: sh /tmp/install.sh --skip-worker-install; plain install fails at worker install stage with 'staged worker did not pass protocol verification'
  • README: installer verifies SHA256 checksum
    In the test: release tarball (~270MB) downloaded and verified by installer at install time; binary content cannot be read beforehand

How to read a tool card

Stitching
pickfit installed this tool itself. Everything under the stitches comes from that run. A tool pickfit has not run never gets stitches.
Verified on your OS
The install steps ran on the system you chose. “Ran on Linux only” means it has not been run on yours yet.
macOS, Linux and WindowsWhere it ran
Green OS marks: pickfit ran the install steps there. Grey ones: the README says it runs there, and pickfit has not tried.
How far the run got
Installs and starts, its main function works, its service answers, it produces model output. A window means a desktop app opened; a plug, an MCP client used it. The mark sits beside the run’s result, under “Why it fits and how we know” and in “How pickfit tested it”.
Visual check
pickfit looked at screenshots of the app. Green: it looks right. Amber with a number: things that look wrong; open it to see them. Grey and dashed: not checked yet.
Known vulnerabilities
Known high or critical vulnerabilities (OSV.dev) in what it installs. Green: none known. Red with a number: how many; open it for the list, and read it before you install. Grey and dashed: not checked, which does not mean safe.
About the run
The day it ran, and the version that ran.
Source of a reason
Every reason a pick fits rests on one profile field. Open the reason to see the field and the value it relies on.
Hidden gem
High quality, little known.
Quality and popularity
Two scores out of 100. Quality: how well it installed and worked when pickfit tried it, and how good its install docs and upkeep are. Popularity: from stars, downloads and Hacker News.
Features checked
Of the main features the README advertises, how many the run actually exercised (green), and how many failed (red). Grey ones were not checked, and the card says why.
Not verified
A well-known tool Fitz mentions from general knowledge. No run record, no steps, no score.