JOHN STOCKWELL, for 12 years a Central Intelligence Agency operative, says in a new book (and on CBS's "60 Minutes") that the CIA bungled its 1975 Angolan operation and misled Congress about it. The CIA director of that period, William Colby, who's also got a book coming out, and other deny the several charges. Who's right? Allegations of bungling about an operation that failed anyway, when Congress pulled the plug, aren't surprising. Allegations of deceiving Congress on some details, while "muzzling" it by informing it confidentially of others, are more intriguing. But who knows? A conclusive answer would probably require a wider perspective than an ex-operator's, and a more unsparing critique than an ex-director's. The Senate Intelligence Committee is investigating - secretly. The CIA books roll on.

What interest us more today is the way Mr. Stockwell entered the fray. Twelve years' exposure to CIA "corruption and mismanagement," he explains, led him to break the secrecy oath he'd signed "as a young person" and to divulge secrets. If the CIA has "kept its act so clean that it was totally above reproach," it would not have had to be "kept honest by . . . public exposure." Courageous whistle blower? Self-indulgent security violator? We lack the basis to judge Mr. Stockwell's psychology and character. But we are troubled by his procedure. He signed an oath, accepted one covert assignment afte another, and broke a trust. One does not have to claim that the secrets he divulges are of the utmost significance to question the successive examples of individual whistle-blowing, some years after the fact, when the CIA has already been put through the wringer, when many reforms have already been instituted and oathers are being proposed, and when it is far too late, in any case, to influence the events being disclosed.

The CIA is testing, in the Snepp case, whether the courts will uphold the agency's secrecy oath, which requires employees to clear manuscripts prior to publication. Regardless, we don't much like oaths as enforcers. They are not statutes, after all, but merely conditions of employment, whether or not it's classified or important, whether or not disclosure would help the country or hurt the agency.

Much better, we think, is William Colby's proposal to end all existing restraints in favor of a simple new law. It would publish an ex-employee's disclosure of intelligence sources and methods "vulnerable to termination or frustration by a foreign power if disclosed." There would be no agency censorship, no prior restraint, no injunctions, no bar or whistle-blowing, no wholesale muzzling. The idea of the law would be to keep an irreducible minimum of legitimate secrets secret. On that ground and perhaps only on that ground, practically everyone could stand.