Merge branch 'feature/exec-pow-files'

Closes: #95 and two stones

Co-authored-by: Roberto Abdelkader Martínez Pérez <robertomartinezp@gmail.com>
This commit is contained in:
pancho horrillo
2020-12-24 14:18:38 +01:00
41 changed files with 268 additions and 172 deletions
-6
View File
@@ -48,11 +48,5 @@ acceptance: install
deps:
@echo "deps here"
docker: build
cp $(BUILD_DIR)/$(BINARY_NAME) $(DOCKER_DIR)/
cp $(DOCS_DIR)/*.pow $(DOCKER_DIR)/
cd $(DOCKER_DIR) && docker build -t kapow .
cd ..
clean:
rm -rf $(BUILD_DIR) $(OUTPUT_DIR) $(DOCKER_DIR)/*
+8 -5
View File
@@ -21,19 +21,22 @@ us easily **turn that into an HTTP API**.
### Let's see this with an example
We want to expose **log entries** for files not found on our **Apache Web
Server**, as an HTTP API. With *Kapow!* we just need to write this file:
Server**, as an HTTP API. With *Kapow!* we just need to write this
*executable* script:
```bash
[apache-host]$ cat search-apache-errors.pow
``` console
[apache-host]$ cat search-apache-errors
#!/usr/bin/env sh
kapow route add /apache-errors - <<-'EOF'
cat /var/log/apache2/access.log | grep 'File does not exist' | kapow set /response/body
cat /var/log/apache2/access.log | grep 'File does not exist' | kapow set /response/body
EOF
[apache-host]$ chmod +x search-apache-errors
```
and then, run it using *Kapow!*
```bash
[apache-host]$ kapow server --bind 0.0.0.0:8080 search-apache-errors.pow
[apache-host]$ kapow server --bind 0.0.0.0:8080 search-apache-errors
```
finally, we can read from the just-defined endpoint:
@@ -11,7 +11,8 @@ In this example we'll be adding the header ``X-Content-Type-Options`` to the res
.. code-block:: console
:linenos:
$ cat sniff.pow
$ cat sniff-route
#!/usr/bin/env sh
kapow route add /sec-hello-world - <<-'EOF'
kapow set /response/headers/X-Content-Type-Options nosniff
kapow set /response/headers/Content-Type text/plain
@@ -19,7 +20,7 @@ In this example we'll be adding the header ``X-Content-Type-Options`` to the res
echo this will be interpreted as plain text | kapow set /response/body
EOF
$ kapow server nosniff.pow
$ kapow server nosniff-route
Testing with :program:`curl`:
@@ -67,7 +68,8 @@ Uploading a file using *Kapow!* is very simple:
.. code-block:: console
:linenos:
$ cat upload.pow
$ cat upload-route
#!/usr/bin/env sh
kapow route add -X POST /upload-file - <<-'EOF'
kapow get /request/files/data/content | kapow set /response/body
EOF
@@ -89,7 +91,8 @@ In this example we reply the line count of the file received in the request:
.. code-block:: console
:linenos:
$ cat count-file-lines.pow
$ cat count-file-lines
#!/usr/bin/env sh
kapow route add -X POST /count-file-lines - <<-'EOF'
# Get sent file
@@ -121,7 +124,8 @@ You can specify custom status code for `HTTP` response:
.. code-block:: console
:linenos:
$ cat error.pow
$ cat error-route
#!/usr/bin/env sh
kapow route add /error - <<-'EOF'
kapow set /response/status 401
echo -n '401 error' | kapow set /response/body
@@ -158,7 +162,8 @@ In this example we'll redirect our users to `Google`:
.. code-block:: console
:linenos:
$ cat redirect.pow
$ cat redirect
#!/usr/bin/env sh
kapow route add /redirect - <<-'EOF'
kapow set /response/headers/Location https://google.com
kapow set /response/status 301
@@ -196,7 +201,8 @@ In the next example we'll set a cookie:
.. code-block:: console
:linenos:
$ cat cookie.pow
$ cat cookie
#!/usr/bin/env sh
kapow route add /setcookie - <<-'EOF'
CURRENT_STATUS=$(kapow get /request/cookies/kapow-status)
+2 -2
View File
@@ -33,7 +33,7 @@ command line:
.. code-block:: console
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile foobar.pow
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile foobar-route
Now *Kapow!* is listening on its default port (8080) accepting requests over
HTTPS. You can test it with the following command:
@@ -60,7 +60,7 @@ CA certificate issuing the client certificates we want to accept with the
.. code-block:: console
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile --clientauth=true --clientcafile path/to/clientCAfile foobar.pow
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile --clientauth=true --clientcafile path/to/clientCAfile foobar-route
With this configuration *Kapow!* will reject connections that do not present a
client certificate or one certificate not issued by the specified CA. You can
+5 -3
View File
@@ -10,7 +10,8 @@ from query params:
.. code-block:: console
:linenos:
$ cat parallel.pow
$ cat parallel-route
#!/usr/bin/env sh
kapow route add '/parallel/{ip1}/{ip2}' - <<-'EOF'
ping -c 1 -- "$(kapow get /request/matches/ip1)" | kapow set /response/body &
ping -c 1 -- "$(kapow get /request/matches/ip2)" | kapow set /response/body &
@@ -30,5 +31,6 @@ Script debugging
Bash provides the ``set -x`` builtin command that "After expanding each simple command,
for command, case command, select command, or arithmetic for command, display the
expanded value of PS4, followed by the command and its expanded arguments or associated
word list". This feature can be used to help debugging the `.pow` scripts and, together
the ``--debug`` option in the server sub-command, the scripts executed in user requests.
word list". This feature can be used to help debugging the init programs and,
together the ``--debug`` option in the server sub-command, the scripts executed
in user requests.
+1 -1
View File
@@ -3,7 +3,7 @@ Examples
.. toctree::
working_with_pow_files
working_with_init_programs
managing_routes
handling_http_requests
using_json
+5 -3
View File
@@ -16,12 +16,13 @@ Example #1
++++++++++
In this example our *Kapow!* service will receive a `JSON` value with an incorrect
date, then our ``pow`` file will fix it and return the correct value to the user.
date, then our init program will fix it and return the correct value to the user.
.. code-block:: console
:linenos:
$ cat fix_date.pow
$ cat fix_date
#!/usr/bin/env sh
kapow route add -X POST /fix-date - <<-'EOF'
kapow set /response/headers/Content-Type application/json
kapow get /request/body | jq --arg newdate "$(date +'%Y-%m-%d_%H-%M-%S')" '.incorrectDate=$newdate' | kapow set /response/body
@@ -46,7 +47,8 @@ order to generate a two-attribute `JSON` response.
.. code-block:: console
$ cat echo-attribute.pow
$ cat echo-attribute
#!/usr/bin/env sh
kapow route add -X POST /echo-attribute - <<-'EOF'
JSON_WHO=$(kapow get /request/body | jq -r .name)
@@ -1,27 +1,28 @@
Working with pow Files
======================
Working with Init Scripts
=========================
Starting *Kapow!* using a pow file
----------------------------------
Starting *Kapow!* using an init script
--------------------------------------
A :file:`pow` file is just a :command:`bash` script, where you make calls to the
``kapow route`` command.
An init program, which can be just a shell script, allows you to make calls to
the ``kapow route`` command.
.. code-block:: console
:linenos:
$ kapow server example.pow
$ kapow server example-init-program
With the :file:`example.pow`:
With the :file:`example-init-program`:
.. code-block:: console
:linenos:
$ cat example.pow
$ cat example-init-program
#!/usr/bin/env sh
#
# This is a simple example of a pow file
# This is a simple example of an init program
#
echo '[*] Starting my script'
echo '[*] Starting my init program'
# We add 2 Kapow! routes
kapow route add /my/route -c 'echo hello world | kapow set /response/body'
@@ -29,32 +30,19 @@ With the :file:`example.pow`:
.. note::
*Kapow!* can be fully configured using just :file:`pow` files
*Kapow!* can be fully configured using just init scripts
Load More Than One pow File
---------------------------
Writing Multiline Routes
------------------------
You can load more than one :file:`pow` file at time. This can help you keep
your :file:`pow` files tidy.
If you need to write more complex actions, you can leverage multiline routes:
.. code-block:: console
:linenos:
$ ls pow-files/
example-1.pow example-2.pow
$ kapow server <(cat pow-files/*.pow)
Writing Multiline pow Files
---------------------------
If you need to write more complex actions, you can leverage multiline commands:
.. code-block:: console
:linenos:
$ cat multiline.pow
$ cat multiline-route
#!/usr/bin/env sh
kapow route add /log_and_stuff - <<-'EOF'
echo this is a quite long sentence and other stuff | tee log.txt | kapow set /response/body
cat log.txt | kapow set /response/body
@@ -77,43 +65,45 @@ Keeping Things Tidy
Sometimes things grow, and keeping things tidy is the only way to mantain the
whole thing.
You can distribute your endpoints in several pow files. And you can keep the
whole thing documented in one html file, served with *Kapow!*.
You can distribute your endpoints in several init programs. And you can keep
the whole thing documented in one html file, served with *Kapow!*.
.. code-block:: console
:linenos:
$ cat index.pow
$ cat index-route
#!/usr/bin/env sh
kapow route add / - <<-'EOF'
cat howto.html | kapow set /response/body
EOF
source ./info_stuff.pow
source ./other_endpoints.pow
source ./info_stuff
source ./other_endpoints
As you can see, the `pow` files can be imported into another `pow` file using
source. In fact, a `pow` file is just a regular shell script.
You can import other shell script libraries with `source`.
Debugging scripts
-----------------
Since *Kapow!* redirects the standard output and the standard error of the `pow`
file given on server startup to its own, you can leverage ``set -x`` to see the
commands that are being executed, and use that for debugging.
Debugging Init Programs/Scripts
-------------------------------
Since *Kapow!* redirects the standard output and the standard error of the init
program given on server startup to its own, you can leverage ``set -x`` to see
the commands that are being executed, and use that for debugging.
To support debugging user request executions, the server subcommand has a
``--debug`` option flag that prompts *Kapow!* to redirect both the script's
standard output and standard error to *Kapow!*'s standard output, so you can
leverage ``set -x`` the same way as with `pow` files.
leverage ``set -x`` the same way as with init programs.
.. code-block:: console
$ cat withdebug.pow
$ cat withdebug-route
#!/usr/bin/env sh
kapow route add / - <<-'EOF'
set -x
echo "This will be seen in the log"
echo "Hi HTTP" | kapow set /response/body
EOF
$ kapow server --debug withdebug.pow
$ kapow server --debug withdebug-route
+1 -1
View File
@@ -75,7 +75,7 @@ Contents
:maxdepth: 2
:caption: Usage Examples
examples/working_with_pow_files
examples/working_with_init_programs
examples/managing_routes
examples/handling_http_requests
examples/using_json
@@ -111,7 +111,7 @@ After building the image you can run the container with:
.. code-block:: console
$ docker run --rm -i -p 8080:8080 -v $(pwd)/whatever.pow:/opt/whatever.pow kapow:latest server /opt/whatever.pow
$ docker run --rm -i -p 8080:8080 -v $(pwd)/whatever-route:/opt/whatever-route kapow:latest server /opt/whatever-route
With the ``-v`` parameter we map a local file into the container's filesystem so
we can use it to configure our *Kapow!* server on startup.
+12 -7
View File
@@ -116,10 +116,10 @@ Install *Kapow!*
Follow the :ref:`installation instructions <installation>`.
Write a :file:`ping.pow` File
+++++++++++++++++++++++++++++
Write an Init Program :file:`ping-route`
+++++++++++++++++++++++++++++++++++++++
*Kapow!* uses plain text files (called `pow` files) where the endpoints you want
*Kapow!* uses init programs/scripts where the endpoints you want
to expose are defined.
For each endpoint, you can decide which commands get executed.
@@ -128,11 +128,15 @@ For our example we need a file like this:
.. code-block:: console
$ cat ping.pow
$ chmod +x ping-route
$ cat ping-route
#!/usr/bin/env sh
kapow route add /ping -c 'ping -c 1 10.10.10.100 | kapow set /response/body'
Let's dissect this beast piece by piece:
#. ``#!/usr/bin/env sh`` - shebang line so that the kernel knows which
interpreter to use
#. ``kapow route add /ping`` - adds a new `HTTP API` endpoint at ``/ping``
path in the *Kapow!* server. You have to use the ``GET`` method to invoke
the endpoint.
@@ -147,18 +151,19 @@ Let's dissect this beast piece by piece:
Launch the Service
++++++++++++++++++
At this point, we only need to launch :program:`kapow` with our :file:`ping.pow`:
At this point, we only need to launch :program:`kapow` with our
:file:`ping-route`:
.. code-block:: console
$ kapow server ping.pow
$ kapow server ping-route
*Kapow!* can expose the user interface through HTTPS, to do this provide the
corresponding key and certificates chain paths at startup:
.. code-block:: console
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile ping.pow
$ kapow server --keyfile path/to/keyfile --certfile path/to/certfile ping-route
Consume the Service
+4 -2
View File
@@ -28,7 +28,8 @@ In this example, an attacker can inject arbitrary parameters to :command:`ls`.
.. code-block:: console
:linenos:
$ cat command-injection.pow
$ cat command-injection
#!/usr/bin/env sh
kapow route add '/vulnerable/{value}' - <<-'EOF'
ls $(kapow get /request/matches/value) | kapow set /response/body
EOF
@@ -48,7 +49,8 @@ request:
.. code-block:: console
:linenos:
$ cat command-injection.pow
$ cat command-injection
#!/usr/bin/env sh
kapow route add '/not-vulnerable/{value}' - <<-'EOF'
ls -- "$(kapow get /request/matches/value)" | kapow set /response/body
EOF
+8 -8
View File
@@ -138,22 +138,22 @@ Let's Backup that Database!
**Senior**
Not at all. The creators of *Kapow!* have thought of everything. You can put
all your route definitions in a special script file and pass it to the server
on startup. They call those files :file:`pow` files and they have
:file:`.pow` extension.
all your route definitions on init programs, which can be shell scripts, and
pass them to the server on startup.
It should look something like:
.. code-block:: console
$ cat backup.pow
$ cat backup-route
#!/usr/bin/env sh
kapow route add -X PUT /db/backup -e ./backup_db.sh
And then you can start *Kapow!* with it:
.. code-block:: console
$ kapow server backup.pow
$ kapow server backup-route
**Junior**
@@ -161,10 +161,10 @@ Let's Backup that Database!
.. code-block:: console
$ kapow server backup.pow
2019/11/26 11:40:01 Running powfile: "backup.pow"
$ kapow server backup-route
2019/11/26 11:40:01 Running init program: "backup-route"
{"id":"19bb4ac7-1039-11ea-aa00-106530610c4d","method":"PUT","url_pattern":"/db/backup","entrypoint":"./backup_db.sh","command":"","index":0}
2019/11/26 11:40:01 Done running powfile: "backup.pow"
2019/11/26 11:40:01 Done running init program: "backup-route"
I understand that this is proof that we have the endpoint available.
+2 -1
View File
@@ -37,10 +37,11 @@ What have we done?
**Junior**
Let me try add this to our :file:`pow` file:
Let me try add this to our init program:
.. code-block:: console
#!/usr/bin/env sh
kapow route add /db/backup_logs -c 'cat /tmp/backup_db.log | kapow set /response/body'
**Senior**
+8 -2
View File
@@ -44,7 +44,9 @@ Securing the server
.. code-block:: console
$ kapow server --keyfile /etc/kapow/tls/keyfile --certfile /etc/kapow/tls/certfile /etc/kapow/awesome.pow
$ kapow server --keyfile /etc/kapow/tls/keyfile \
--certfile /etc/kapow/tls/certfile \
/etc/kapow/awesome-route
It's easy, please copy the private key file and certificate chain to `/etc/kapow/tls` and we can restart.
@@ -70,7 +72,11 @@ Securing the server
.. code-block:: console
$ kapow server --keyfile /etc/kapow/tls/keyfile --certfile /etc/kapow/tls/certfile --clientauth=true --clientcafile /etc/kapow/tls/clientCAfile /etc/kapow/awesome.pow
$ kapow server --keyfile /etc/kapow/tls/keyfile \
--certfile /etc/kapow/tls/certfile \
--clientauth true \
--clientcafile /etc/kapow/tls/clientCAfile \
/etc/kapow/awesome-route
Done!
@@ -5,7 +5,7 @@ A small web gui for [pandoc](https://pandoc.org) that allows to convert between
## How to run it
```
$ kapow server DocumentConverter.pow
$ kapow server DocumentConverter
```
@@ -10,7 +10,7 @@ Run a long network scan in background with support for webhook on completion.
## How to run it
```
$ kapow server NetworkScanner.pow
$ kapow server NetworkScanner
```
@@ -1,3 +1,3 @@
#!/bin/sh
#!/usr/bin/env sh
kapow route add /sniff -c 'tcpdump -i any -U -s0 -w - "not portrange 8080-8082" | kapow set /response/body'
@@ -8,7 +8,7 @@ Provides an HTTP service that allows the user to sniff the network in real time.
For the sake of simplicity, run:
```
$ sudo -E kapow server NetworkSniffer.pow
$ sudo -E kapow server NetworkSniffer
```
In a production environment, tcpdump should be run with the appropiate
+1 -1
View File
@@ -6,7 +6,7 @@ A simple "Hello World!" type example.
## How to run it
```
$ kapow server HelloWorld.pow
$ kapow server HelloWorld
```
@@ -1,4 +1,4 @@
#!/bin/sh
#!/usr/bin/env sh
kapow route add /apache-errors - <<-'EOF'
cat /var/log/apache2/access.log | grep 'File does not exist' | kapow set /response/body
+1 -1
View File
@@ -5,7 +5,7 @@ A simple service that exposes log entries for files not found on our Apache Web
## How to run it
```
$ kapow server FixLogGrep.pow
$ kapow server FixLogGrep
```
+1 -1
View File
@@ -5,7 +5,7 @@ A simple service that exposes log entries that matches **with a given text** on
## How to run it
```
$ kapow server DynamicLogGrep.pow
$ kapow server DynamicLogGrep
```
+2 -2
View File
@@ -5,7 +5,7 @@ Expose several system properties and logs.
## How to run it
```
$ kapow server SystemMonitor.pow
$ kapow server SystemMonitor
```
@@ -39,7 +39,7 @@ drwxr-xr-x 15 root root 4096 Jan 27 07:34 ..
```
$ curl -s http://localhost:8080/process
nil 46717 0.0 0.0 111224 8196 pts/2 Sl 16:48 0:00 kapow server SystemMonitor.pow
nil 46717 0.0 0.0 111224 8196 pts/2 Sl 16:48 0:00 kapow server SystemMonitor
root 47405 0.0 0.0 0 0 ? I 16:50 0:00 [kworker/3:1-mm_percpu_wq]
root 47406 0.0 0.0 0 0 ? I 16:50 0:00 [kworker/0:1]
root 47819 0.0 0.0 0 0 ? I 16:52 0:00 [kworker/7:2-mm_percpu_wq]
+7 -3
View File
@@ -27,10 +27,14 @@ import (
// AddRoute will add a new route in kapow
func AddRoute(host, path, method, entrypoint, command string, w io.Writer) error {
url := host + "/routes"
body, _ := json.Marshal(map[string]string{
payload := map[string]string{
"method": method,
"url_pattern": path,
"entrypoint": entrypoint,
"command": command})
"command": command,
}
if entrypoint != "" {
payload["entrypoint"] = entrypoint
}
body, _ := json.Marshal(payload)
return http.Post(url, "application/json", bytes.NewReader(body), w)
}
-1
View File
@@ -31,7 +31,6 @@ func TestSuccessOnCorrectRoute(t *testing.T) {
JSON(map[string]string{
"method": "GET",
"url_pattern": "/hello",
"entrypoint": "",
"command": "echo Hello World | kapow set /response/body",
}).
Reply(http.StatusCreated).
+1 -1
View File
@@ -80,7 +80,7 @@ func init() {
// TODO: Add default values for flags and remove path flag
routeAddCmd.Flags().String("control-url", getEnv("KAPOW_CONTROL_URL", "http://localhost:8081"), "Kapow! control interface URL")
routeAddCmd.Flags().StringP("method", "X", "GET", "HTTP method to accept")
routeAddCmd.Flags().StringP("entrypoint", "e", "/bin/sh -c", "Command to execute")
routeAddCmd.Flags().StringP("entrypoint", "e", "", "Command to execute")
routeAddCmd.Flags().StringP("command", "c", "", "Command to pass to the shell")
var routeRemoveCmd = &cobra.Command{
+11
View File
@@ -0,0 +1,11 @@
// +build !windows
package cmd
import (
"os/exec"
)
func BuildCmd(path string) *exec.Cmd {
return exec.Command(path)
}
+9
View File
@@ -0,0 +1,9 @@
package cmd
import (
"os/exec"
)
func BuildCmd(path string) *exec.Cmd {
return exec.Command("cmd.exe", "/c", path)
}
+48 -19
View File
@@ -17,9 +17,11 @@
package cmd
import (
"bufio"
"errors"
"io"
"os"
"os/exec"
"sync"
"github.com/spf13/cobra"
@@ -29,7 +31,7 @@ import (
// ServerCmd is the command line interface for kapow server
var ServerCmd = &cobra.Command{
Use: "server [optional flags] [optional pow file(s)]",
Use: "server [optional flags] [optional init program(s)]",
Short: "Start a kapow server",
Long: `Start a Kapow server with a client interface, a data interface and an
admin interface`,
@@ -57,23 +59,8 @@ var ServerCmd = &cobra.Command{
server.StartServer(sConf)
if len(args) > 0 {
powfile := args[0]
_, err := os.Stat(powfile)
if os.IsNotExist(err) {
logger.L.Fatalf("%s does not exist", powfile)
}
logger.L.Printf("Running powfile: %q\n", powfile)
kapowCMD := exec.Command("bash", powfile)
kapowCMD.Stdout = os.Stdout
kapowCMD.Stderr = os.Stderr
kapowCMD.Env = os.Environ()
err = kapowCMD.Run()
if err != nil {
logger.L.Fatal(err)
}
logger.L.Printf("Done running powfile: %q\n", powfile)
for _, path := range args {
go Run(path, sConf.Debug)
}
select {}
@@ -112,3 +99,45 @@ func validateServerCommandArguments(cmd *cobra.Command, args []string) error {
return nil
}
func Run(path string, debug bool) {
logger.L.Printf("Running init program %+q", path)
cmd := BuildCmd(path)
cmd.Env = os.Environ()
var wg sync.WaitGroup
if debug {
if stdout, err := cmd.StdoutPipe(); err == nil {
wg.Add(1)
go logPipe(path, "stdout", stdout, &wg)
}
if stderr, err := cmd.StderrPipe(); err == nil {
wg.Add(1)
go logPipe(path, "stderr", stderr, &wg)
}
}
err := cmd.Start()
if err != nil {
logger.L.Fatalf("Unable to run init program %+q: %s", path, err)
}
wg.Wait()
err = cmd.Wait()
if err != nil {
logger.L.Printf("Init program exited with error: %s", err)
} else {
logger.L.Printf("Init program %+q finished OK", path)
}
}
func logPipe(path, name string, pipe io.ReadCloser, wg *sync.WaitGroup) {
defer wg.Done()
in := bufio.NewScanner(pipe)
for in.Scan() {
logger.L.Printf("%+q (%s): %s", path, name, in.Text())
}
if err := in.Err(); err != nil {
logger.L.Printf("Error reading from %+qs %s: %s", path, name, err)
}
}
+4
View File
@@ -132,6 +132,10 @@ func addRoute(res http.ResponseWriter, req *http.Request) {
return
}
if route.Entrypoint == "" {
route.Entrypoint = defaultEntrypoint
}
route.ID = id.String()
created := funcAdd(route)
+5
View File
@@ -0,0 +1,5 @@
// +build !windows
package control
var defaultEntrypoint = "/bin/sh -c"
@@ -0,0 +1,3 @@
package control
var defaultEntrypoint = "cmd.exe /c"
+3 -3
View File
@@ -19,7 +19,7 @@ package model
// Route contains the data needed to represent a Kapow! user route.
type Route struct {
// ID is the unique identifier of the Route.
ID string `json:"id"`
ID string `json:"id,omitempty"`
// Method is the HTTP method that will match this Route.
Method string `json:"method"`
@@ -33,7 +33,7 @@ type Route struct {
//
// This string will be split according to the shell parsing rules to
// be passed as a list to exec.Command.
Entrypoint string `json:"entrypoint"`
Entrypoint string `json:"entrypoint,omitempty"`
// Command is the last argument to be passed to exec.Command when
// executing the Entrypoint
@@ -43,5 +43,5 @@ type Route struct {
// It is an output field, its value is ignored as input.
Index int `json:"index"`
Debug bool `json:"debug"`
Debug bool `json:"debug,omitempty"`
}
+43 -41
View File
@@ -219,7 +219,7 @@ field must be an escaped JSON string.
}
]
```
* **Sample Call**: `$ curl $KAPOW_URL/routes`
* **Sample Call**: `$ curl $KAPOW_DATA_URL/routes`
* **Notes**: Currently all routes are returned; in the future, a filter may be
accepted.
@@ -260,7 +260,7 @@ A new id is created for the appended route so it can be referenced later.
* **Code**: `422`; **Reason**: `Invalid Route`
* **Sample Call**:<br />
```sh
$ curl -X POST --data-binary @- $KAPOW_URL/routes <<EOF
$ curl -X POST --data-binary @- $KAPOW_DATA_URL/routes <<EOF
{
"method": "GET",
"url_pattern": "/hello",
@@ -313,7 +313,7 @@ A new id is created for the appended route so it can be referenced later.
* **Code**: `422`; Reason: `Invalid Route`
* **Sample Call**:<br />
```sh
$ curl -X PUT --data-binary @- $KAPOW_URL/routes <<EOF`
$ curl -X PUT --data-binary @- $KAPOW_DATA_URL/routes <<EOF`
{
"method": "GET",
"url_pattern": "/hello",
@@ -346,7 +346,7 @@ Removes the route identified by `{id}`.
* **Code**: `404`; Reason: `Route Not Found`
* **Sample Call**:<br />
```sh
$ curl -X DELETE $KAPOW_URL/routes/ROUTE_1f186c92_f906_4506_9788_a1f541b11d0f
$ curl -X DELETE $KAPOW_DATA_URL/routes/ROUTE_1f186c92_f906_4506_9788_a1f541b11d0f
```
* **Notes**:
@@ -374,7 +374,7 @@ Retrieves the information about the route identified by `{id}`.
* **Code**: `404`; Reason: `Route Not Found`
* **Sample Call**:<br />
```sh
$ curl -X GET $KAPOW_URL/routes/ROUTE_1f186c92_f906_4506_9788_a1f541b11d0f
$ curl -X GET $KAPOW_DATA_URL/routes/ROUTE_1f186c92_f906_4506_9788_a1f541b11d0f
```
* **Notes**:
@@ -610,29 +610,30 @@ Commands:
This command runs the Kapow! server, which is the core of Kapow!. If
run without parameters, it will run an unconfigured server. It can accept a path
to a `pow` file, which is a shell script that contains commands to configure
the Kapow! server.
to an executable file, the init program, which can be a shell script that
contains commands to configure the *Kapow!* server.
The `pow` can leverage the `kapow route` command, which is used to define a route.
The `kapow route` command needs a way to reach the Kapow! server, and for that,
`kapow` provides the `KAPOW_URL` variable in the environment of the
aforementioned shell script.
The init program can leverage the `kapow route` command, which is used to define
a route. The `kapow route` command needs a way to reach the *Kapow!* server,
and for that, `kapow` provides the `KAPOW_DATA_URL` variable in the environment
of the aforementioned init program.
Every time the kapow server receives a request, it will spawn a process to
handle it, according to the specified entrypoint, `/bin/sh -c` by default, and then
execute the specified command. This command is tasked with processing the
incoming request, and can leverage the `request` and `response` commands to
easily access the `HTTP Request` and `HTTP Response`, respectively.
Every time the *Kapow!* server receives a request, it will spawn a process to
handle it, according to the specified entrypoint, `/bin/sh -c` by default in
unices, and `cmd.exe /c` in Windows®, and then execute the specified command.
This command is tasked with processing the incoming request, and can leverage
the `get` and `set` commands to easily access the `HTTP Request` and `HTTP
Response`, respectively.
In order for `request` and `response` to do their job, they require a way to
reach the Kapow! server, as well as a way to identify the current request being
served. Thus, the Kapow! server adds the `KAPOW_URL` and `KAPOW_HANDLER_ID` to the
In order for `get` and `set` to do their job, they require a way to reach the
*Kapow!* server, as well as a way to identify the current request being served.
Thus, the *Kapow!* server adds the `KAPOW_DATA_URL` and `KAPOW_HANDLER_ID` to the
process' environment.
#### Example
```sh
$ kapow server /path/to/service.pow
``` console
$ kapow server /path/to/service
```
@@ -654,11 +655,11 @@ To deregister a route you must provide a *route_id*.
#### **Environment**
- `KAPOW_URL`
- `KAPOW_DATA_URL`
#### **Help**
```sh
``` console
$ kapow route --help
Usage: kapow route [OPTIONS] COMMAND [ARGS]...
@@ -669,7 +670,7 @@ Commands:
add
remove
```
```sh
``` console
$ kapow route add --help
Usage: kapow route add [OPTIONS] URL_PATTERN [COMMAND_FILE]
@@ -680,7 +681,7 @@ Options:
--url TEXT
--help Show this message and exit.
```
```sh
``` console
$ kapow route remove --help
Usage: kapow route remove [OPTIONS] ROUTE_ID
@@ -691,8 +692,8 @@ Options:
#### Example
```sh
kapow route add -X GET '/list/{ip}' -c 'nmap -sL $(kapow get /request/matches/ip) | kapow set /response/body'
``` console
$ kapow route add -X GET '/list/{ip}' -c 'nmap -sL $(kapow get /request/matches/ip) | kapow set /response/body'
```
### `request`
@@ -701,14 +702,14 @@ Exposes the requests' resources.
#### **Environment**
- `KAPOW_URL`
- `KAPOW_DATA_URL`
- `KAPOW_HANDLER_ID`
#### Example
```sh
# Access the body of the request
kapow get /request/body
``` console
$ # Access the body of the request
$ kapow get /request/body
```
@@ -718,27 +719,28 @@ Exposes the response's resources.
#### **Environment**
- `KAPOW_URL`
- `KAPOW_DATA_URL`
- `KAPOW_HANDLER_ID`
#### Example
```sh
# Write to the body of the response
echo 'Hello, World!' | kapow set /response/body
``` console
$ # Write to the body of the response
$ echo 'Hello, World!' | kapow set /response/body
```
## An End-to-End Example
```sh
$ cat nmap.kpow
``` console
$ cat nmap-route
#!/usr/bin/env sh
kapow route add -X GET '/list/{ip}' -c 'nmap -sL $(kapow get /request/matches/ip) | kapow set /response/body'
```
```sh
$ kapow ./nmap.kapow
``` console
$ kapow server ./nmap-route
```
```sh
$ curl $KAPOW_URL/list/127.0.0.1
``` console
$ curl $KAPOW_DATA_URL/list/127.0.0.1
Starting Nmap 7.70 ( https://nmap.org ) at 2019-05-30 14:45 CEST
Nmap scan report for localhost (127.0.0.1)
Host is up (0.00011s latency).
+19
View File
@@ -0,0 +1,19 @@
package main
import (
"fmt"
"os"
"os/exec"
)
func main() {
cmd := exec.Command("rundll32.exe", "url.dll,FileProtocolHandler", os.Args[1])
err := cmd.Start()
if err != nil {
fmt.Println(err)
}
err = cmd.Wait()
if err != nil {
fmt.Println(err)
}
}