WHC
2fde898da0
Tools - Table-driven tool registry with permission classes and default/minimal profiles; new tools: edit_file, list_dir, glob, grep, delete_file, move_file, todo_write, web_fetch, ask_user (plus read_file paging). - Built-in Myers diff; write tools return unified diffs. - Persistent permission rules (.agent/permissions.json), per-turn undo snapshots (.agent/undo), AGENTS.md project notes, auto-compact at 80% of the model context, cancellable run_shell with process-group kill. TUI - Table-driven slash commands: /provider /key /effort /status /tools /permissions /yolo /undo /retry /diff /export /copy /rename /delete /init /max-steps /todo, with / and @ completion. - Permission modal with diff and prefix rules, ask_user modal, todo panel, header with model/effort/context, bracketed paste, line editing, wide-char cursor, NO_COLOR and ASCII fallback; Esc never quits. Startup and configuration - API keys are entered in the TUI (hidden) and saved per provider in <config>/credentials (0600); env variables are only a fallback. - Last provider, per-provider model and reasoning effort are remembered in <config>/settings.json; --effort/-e and /effort send reasoning_effort (low|medium|high|max) to models that accept it. - DeepSeek preset uses deepseek-flash / deepseek-v4-pro. - Session flags --new/--continue/--resume/--list-sessions, --yolo, --confirm. Compatibility and tests - ncurses detection, macOS/Homebrew support, make install, XDG config dir, POSIX cleanups; Gitea CI (gcc/clang/asan/musl/macOS). - Unit tests for diff, fs tools, registry, undo/perms and config; e2e fixtures multi_file_refactor and ask_or_todo; vendored cJSON.
95 lines
3.2 KiB
C
95 lines
3.2 KiB
C
#define _POSIX_C_SOURCE 200809L
|
|
|
|
#include "cst_agent.h"
|
|
#include "os_compat.h"
|
|
|
|
#include <stdio.h>
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
#include <sys/stat.h>
|
|
#include <unistd.h>
|
|
|
|
/* run_shell must never hang the agent: timeout kills the whole process
|
|
* group, stdin is empty, huge output is capped. */
|
|
|
|
static int nfail;
|
|
|
|
#define EXPECT(cond, msg) \
|
|
do { \
|
|
if (!(cond)) { \
|
|
fprintf(stderr, "FAIL %s:%d: %s\n", __FILE__, __LINE__, msg); \
|
|
nfail++; \
|
|
} \
|
|
} while (0)
|
|
|
|
static char g_ws[256];
|
|
|
|
static int sh(const char *cmd, cst_tool_result_t *r, double *ms) {
|
|
ctl_tool_call_t c;
|
|
memset(&c, 0, sizeof(c));
|
|
snprintf(c.id, sizeof(c.id), "s");
|
|
snprintf(c.name, sizeof(c.name), "run_shell");
|
|
snprintf(c.args[0].key, sizeof(c.args[0].key), "cmd");
|
|
c.args[0].value = (char *)cmd;
|
|
c.args[0].value_len = strlen(cmd);
|
|
c.nargs = 1;
|
|
memset(r, 0, sizeof(*r));
|
|
double t0 = os_now_ms();
|
|
cst_tool_dispatch(g_ws, &c, r);
|
|
*ms = os_now_ms() - t0;
|
|
return r->exit_code;
|
|
}
|
|
|
|
int main(void) {
|
|
snprintf(g_ws, sizeof(g_ws), "/tmp/cst-shell-test-%d", (int)getpid());
|
|
mkdir(g_ws, 0755);
|
|
cst_tool_result_t r;
|
|
double ms;
|
|
|
|
EXPECT(sh("echo hi; exit 3", &r, &ms) == 3, "exit code passed through");
|
|
EXPECT(r.out && strstr(r.out, "hi"), "stdout captured");
|
|
cst_tool_result_free(&r);
|
|
|
|
/* stdin is /dev/null: a command reading stdin returns at once */
|
|
sh("cat; echo done", &r, &ms);
|
|
EXPECT(r.out && strstr(r.out, "done") && ms < 5000, "stdin is empty");
|
|
cst_tool_result_free(&r);
|
|
|
|
/* timeout kills the process group, including background children that
|
|
* keep the output pipe open */
|
|
setenv("CST_SHELL_TIMEOUT", "1", 1);
|
|
os_set_run_timeout(1);
|
|
int ec = sh("sleep 30 & sleep 30; echo never", &r, &ms);
|
|
EXPECT(ec == 124, "timeout exit code 124");
|
|
EXPECT(ms < 8000, "timeout returns promptly");
|
|
EXPECT(!r.out || !strstr(r.out, "never"), "command was stopped");
|
|
cst_tool_result_free(&r);
|
|
os_set_run_timeout(0);
|
|
unsetenv("CST_SHELL_TIMEOUT");
|
|
|
|
/* the shell exits but leaves a background child holding the pipe */
|
|
ec = sh("sleep 30 & echo started", &r, &ms);
|
|
EXPECT(ec == 0, "shell exit code kept when a background child lingers");
|
|
EXPECT(r.out && strstr(r.out, "started"), "foreground output kept");
|
|
EXPECT(ms < 5000, "lingering background child does not block");
|
|
cst_tool_result_free(&r);
|
|
|
|
/* huge output: capped and marked, the child is not blocked on the pipe */
|
|
sh("head -c 5000000 /dev/zero | tr '\\0' x; echo; echo tail-marker", &r, &ms);
|
|
EXPECT(r.out && r.out_len < 2000000, "output capped");
|
|
EXPECT(r.out && strstr(r.out, "truncated"), "truncation noted");
|
|
EXPECT(ms < 20000, "large output does not hang");
|
|
cst_tool_result_free(&r);
|
|
|
|
char cmd[300];
|
|
snprintf(cmd, sizeof(cmd), "rm -rf %s", g_ws);
|
|
if (system(cmd) != 0)
|
|
fprintf(stderr, "warning: cleanup failed\n");
|
|
if (nfail) {
|
|
fprintf(stderr, "test_shell: %d failed\n", nfail);
|
|
return 1;
|
|
}
|
|
printf("test_shell: OK\n");
|
|
return 0;
|
|
}
|